Back to GetFilings.com

Table of Contents

UNITED STATES SECURITIES AND EXCHANGE COMMISSION

WASHINGTON, D.C. 20549

FORM 10-K

(MARK ONE)

[ x ] Annual Report Pursuant to Section 13 or 15(d) of the Securities Exchange Act of 1934

For the Fiscal Year Ended December 31, 2004.

or

[  ] Transition Report Pursuant to Section 13 or 15(d) of the Securities Exchange Act of 1934

For the Transition Period from              to             .

Commission File Number: 000-31633

KOSAN BIOSCIENCES INCORPORATED

(Exact name of registrant as specified in its charter)

3832 BAY CENTER PLACE, HAYWARD, CALIFORNIA 94545

(address of principal executive offices)

(510) 732-8400

(Registrant’s telephone number, including area code)

Securities registered pursuant to Section 12(b) of the Act: NONE

Securities registered pursuant to Section 12(g) of the Act:

COMMON STOCK, $.001 PAR VALUE PER SHARE

Indicate by check mark whether the registrant (1) has filed all reports required to be filed by Section 13 or 15(d) of the Securities Exchange Act of 1934 during the preceding 12 months (or for such shorter period that the registrant was required to file such reports), and (2) has been subject to such filing requirements for the past 90 days. Yes [ x ]    No [  ]

Indicate by check mark if disclosure of delinquent filers pursuant to Item 405 of Regulation S-K is not contained herein, and will not be contained, to the best of registrant’s knowledge, in definitive proxy or information statements incorporated by reference in Part III of this Form 10-K or any amendment to this Form 10-K [ x ].

Indicate by check mark whether the registrant is an accelerated filer (as defined in Exchange Act Rule 12b-2). Yes [ x ]    No [  ]

The aggregate market value of the voting stock held by non-affiliates of the registrant was approximately $144,470,792 as of June 30, 2004, based upon the closing sale price on the Nasdaq National Market reported for such date. The determination of affiliate status for the purposes of this calculation is not necessarily a conclusive determination for other purposes. The calculation excludes approximately 10,656,794 shares held by directors, officers and stockholders whose ownership exceeded five percent of the registrant’s outstanding Common Stock as of June 30, 2004. Exclusion of these shares should not be construed to indicate that such person controls, is controlled by or is under common control with the registrant.

The number of shares of common stock of the registrant outstanding at February 28, 2005 was 29,097,833 shares.

DOCUMENTS INCORPORATED BY REFERENCE

Certain portions of the registrant’s Proxy Statement in connection with the registrant’s 2005 Annual Meeting of Stockholders, to be filed with the Securities and Exchange Commission pursuant to Regulation 14A not later than 120 days after the end of the fiscal year covered by this report on Form 10-K, are incorporated by reference into Part III of this report.

Table of Contents

Kosan Biosciences Incorporated

Form 10-K

For the Fiscal Year Ended December 31, 2004

TABLE OF CONTENTS

2

Table of Contents

PART I

Forward-Looking Statements and Risk Factors

This annual report on Form 10-K contains forward-looking statements within the meaning of Section 27A of the Securities Act of 1933, as amended, and within the meaning of Section 21E of the Securities Exchange Act of 1934, as amended, which are subject to the “safe harbor” created by those sections. These forward-looking statements include, but are not limited to, the statements identified in “Item 1. Business” under the heading “Forward-Looking Statements.” In some cases, forward-looking statements can be identified by words such as “expect,” “anticipate,” “intend,” “believe,” “hope,” “assume,” “estimate,” “plan,” “will” and other similar words and expressions. In addition, any statements that refer to expectations, projections or other characterizations of future events or circumstances are forward-looking statements.

These forward-looking statements involve risks and uncertainties, including the risks discussed in “Item 1. Business—Risks Factors That May Affect Results of Operations and Financial Condition,” that could cause our actual results to differ materially from those in the forward-looking statements. We undertake no obligation to publicly release any revisions to the forward-looking statements or reflect events or circumstances after the date of this report. The risks discussed in “Item 1. Business—Risks Factors That May Affect Results of Operations and Financial Condition” and elsewhere in this report should be considered in evaluating our prospects and future financial performance.

ITEM 1. BUSINESS

Overview

We are a biotechnology company developing drug candidates from an important class of natural product compounds known as polyketides. Polyketides are naturally made in very small amounts in certain microorganisms and are difficult to make or modify chemically. Using our expertise and proprietary technologies, we are able to create, modify and produce polyketides in a variety of ways. Because polyketides have been a rich source of marketed drugs, creating novel polyketides and having the capability to more cost-effectively produce polyketides should provide us with a pipeline of potential drug candidates that address existing or potential large markets. We have two first-in-class, polyketide-derived agents in Phase II and Phase Ib clinical trials for the treatment of cancer: KOS-862 (Epothilone D) and 17-AAG, a geldanamycin analog and Hsp90 inhibitor.

Our lead drug candidate, KOS-862, is an epothilone with a mechanism of action similar to taxanes, one of the most successful classes of anti-tumor agents introduced in the last decade. We are co-developing KOS-862, as well as a second-generation epothilone, KOS-1584, with Hoffmann-La Roche, Inc. and F. Hoffmann-La Roche Ltd., collectively Roche, and have U.S. co-promotion rights to any resulting epothilones that receive regulatory approval for marketing. KOS-862 is in Phase II and Phase Ib clinical trials, and KOS-1584 is in a Phase I clinical trial.

The most advanced compound in our Hsp90 (heat shock protein 90) inhibitor program, 17-AAG, has a novel mechanism of action targeting multiple pathways involved in cancer cell growth and survival. We are collaborating with the National Cancer Institute, or NCI, a component of the National Institutes of Health, or NIH, in the development of 17-AAG and other analogs of geldanamycin, a polyketide, for the treatment of cancer. 17-AAG is in Phase II and Phase Ib combination clinical trials being conducted by the NCI. We are testing a proprietary formulation of 17-AAG, KOS-953, in Phase I and Phase Ib clinical trials. In addition, in collaboration with the NCI, we are developing KOS-1022 (DMAG), a second-generation geldanamycin analog, in Phase I clinical trials.

We also have additional polyketide-based product candidates in our pipeline, including compounds for cancer, gastrointestinal motility and infectious disease that are undergoing preclinical evaluation, as well as several early stage product research and technology development programs. These programs seek to alter the structure of high-value polyketide pharmaceuticals to confer advantageous properties, to use heterologous over-expression to improve production and to discover new polyketide structures for important unmet clinical needs.

3

Table of Contents

KOSAN BIOSCIENCES

We were incorporated under the laws of the state of California in January 1995 and commenced operations in 1996. In July 2000, we were reincorporated under the laws of the state of Delaware.

Overview of Polyketides

Polyketides are complex natural products produced by certain microorganisms. There are about 10,000 known polyketides, from which numerous pharmaceutical products in many therapeutic areas have been derived. We focus on polyketides because of their demonstrated ability to address large markets and to treat many different disease conditions. The following table lists some of the important polyketides and their uses.

Selected Polyketide Products and Their Uses

Unlike most classes of compounds, different polyketides often have unrelated structures. The common features that link the polyketides as a class are the sequence of reactions by which they are formed and the intermediate compounds made in these reactions. Each polyketide is produced by a unique polyketide synthase, or PKS, which is a large enzyme composed of many component enzymes. There are two types of PKSs, modular and iterative. We focus primarily on modular PKSs. Modular PKSs contain many enzymes, each of which is used only once during polyketide production, while iterative PKSs may use some enzymes several times. Erythromycin is an example of a polyketide made by a modular PKS. Erythromycin is not only a valuable antibiotic product, but it is also used in the production of the antibiotics azithromycin and clarithromycin.

A modular PKS is composed of multiple proteins. The instructions for making these proteins are contained in the DNA, or genome, of the microorganism that produces the modular PKS. Each protein has its own set of instructions, which is called the gene for the protein. The complete set of genes for a modular PKS that produces a polyketide is called a gene cluster. When any gene of a polyketide gene cluster is identified, the entire cluster can usually be easily obtained. We accomplish this by using recombinant DNA technology, which is a set of techniques that enables scientists to combine or alter DNA from different organisms.

The synthesis of a polyketide essentially involves the linking of a number of small building block compounds to form the larger polyketide. A modular PKS can be functionally subdivided into units called modules, each of which is responsible for a single building block used in the synthesis. Synthesis begins at the first module, called the loading module, located at one end of the PKS, and continues to the end through multiple extender modules, each of which

4

Table of Contents

adds and modifies another building block. This process creates a chain of building blocks that forms the polyketide; in many instances, the chain is linked to itself to form a ring. Because each module codes for one building block, the number of modules in a PKS codes for the size of the polyketide. Each module contains three essential enzyme activities responsible for connecting the polyketide building blocks. One of these activities selects which building block (there are at least four different building blocks) is used, and the other two are involved in linking the building block to the growing chain and passing the chain to the next module. A module may also have one, two or three additional enzymes that modify the building blocks once they are incorporated into the chain.

The structure of a polyketide can therefore be viewed as being determined by the types, order and number of modules in a modular PKS. The types of modules dictate which building blocks are used, the order of the modules dictates the sequence of building blocks and the number of modules dictates the number of building blocks in, or size of, the polyketide. Modifying, adding or deleting modules results in specific and predictable changes to the structure of the polyketide. Polyketides may also be made by sophisticated duplicate synthesis techniques, or by a combination of chemical and biological techniques.

Our Strategy

Our goal is to create a pipeline of drug candidates that can be advanced into clinical trials and developed into commercialized drugs. Our focus is on drug candidates that address existing or potential large markets. Our strategy includes the following components:

Maximize Value, Minimize Risk. Our principal programs are focused on the development of improved versions of known compounds or drugs and increasing the efficiency of large-scale production of polyketides. By improving the properties of known compounds, we believe we can create novel products that take advantage of the known utility, safety, development path and market for existing drugs to reduce the risk and time required for development. By improving the efficiency of large-scale production, we can more cost-effectively access promising compounds that cannot be obtained from nature in sufficient quantities to permit development and commercialization.

Establish Collaborative Relationships. We have a collaboration with Roche to co-develop and to commercialize our anticancer product candidates, KOS-862 and KOS-1584, for the treatment of cancer. We have a collaboration with the NCI to clinically develop 17-AAG (including KOS-953) and KOS-1022 (DMAG) for the treatment of cancer. We plan to establish additional collaborative relationships with pharmaceutical companies and other research and development organizations to help us advance product candidates into or through clinical trials and to the market, obtain additional product candidates and apply our technologies to create new polyketides and develop large-scale production systems.

Acquire Complementary Technologies and Products. We have license agreements with Stanford University in the area of polyketide genetic engineering, the Sloan-Kettering Institute for Cancer Research, or Sloan-Kettering, in the area of epothilones and with the NCI in the area of geldanamycin analogs. We may acquire or license additional complementary technologies or product candidates from others.

Enhance Leadership Position of Our Technology Platform. We intend to expand and enhance our technologies by continuing in-house research activities on polyketide biosynthesis and chemical synthesis. We plan to continue to extend the reach of our technologies through strategic alliances or acquisitions. We plan to broaden and protect our intellectual property portfolio and in-license patents that complement our core technologies and products.

Product Development Opportunities

Our primary programs are currently directed at discovery and development of polyketides for cancer, gastrointestinal motility disorders and infectious disease. These programs were selected because they represent opportunities where our technologies could improve upon existing products or fill unmet needs and because each addresses existing or potential large markets.

5

Table of Contents

KOSAN BIOSCIENCES

Cancer

Epothilones

Epothilones are polyketides that inhibit cancer cells by the same mechanism as a class of cytotoxic drugs known as taxanes, including paclitaxel, marketed as Taxol^® by Bristol-Myers Squibb Company, and docetaxel, marketed as Taxotere^® by Aventis Pharmaceuticals, Inc. In 2004, worldwide net sales of taxanes were approximately $2.8 billion. The epothilones may address a key limitation of the taxane class because they are active against taxane-resistant human tumor cell lines in in vitro and in vivo animal models, as well as being active against taxane-sensitive human tumors cells in these models.

In September 2002, we entered into an agreement with Roche to jointly develop KOS-862, our lead epothilone product candidate, and other epothilones. KOS-862 is being evaluated in Phase II clinical trials in advanced breast and prostate cancers with Roche. These Phase II studies are Simon two-stage design trials that have an interim review and decision point following the completion of stage one before advancing to stage two. We are conducting the prostate cancer trial at multiple centers in North America, and Roche is conducting the breast cancer trial at multiple centers. In addition to the Phase II monotherapy trials, KOS-862 is being evaluated in Phase Ib clinical trials in combination with Gemzar^®, Paraplatin^® and Herceptin^®, three standard-of-care anticancer chemotherapeutics. KOS-1584 is our second-generation Epothilone D anticancer product candidate that we are developing in collaboration with Roche. In December 2004, we initiated a Phase I clinical trial of KOS-1584 in solid tumors. Roche is responsible for the majority of the costs of the epothilone clinical trials conducted under our collaboration and is funding all of the KOS-862 and KOS-1584 clinical trials other than the KOS-862 prostate cancer clinical trial, which we are currently funding. Competing epothilones in clinical development include those being developed by Bristol-Myers Squibb (reported to be in Phase III clinical trials), Novartis AG (reported to be in Phase II clinical trials) and Schering AG (reported to be in Phase I clinical trials). The following table summarizes the current development status of our epothilone program.

Hsp90 Inhibitors

We are developing analogs of the polyketide geldanamycin that have been shown to cause the degradation of numerous proteins involved in the growth and survival of cancer cells. These compounds bind to and disrupt the function of Hsp90 (heat shock protein 90). Hsp90 is a molecular “chaperone” in cells that maintains the stability and function of “client proteins” implicated in signal transduction as well as the proliferation, angiogenesis and metastasis of tumors. By preventing Hsp90 from protecting its “clients,” geldanamycin analogs cause their degradation. By simultaneously depleting multiple proteins involved in the genesis and maintenance of cancer cells, geldanamycin analogs may serve as chemotherapeutic agents in a number of cancers. Moreover, preclinical studies demonstrate that these compounds are synergistic with certain other inhibitors of the signal transduction client proteins, as well as several conventional anticancer agents.

In October 2002, we entered into two Cooperative Research and Development Agreements (CRADAs) with the NIH to collaborate with the NCI on the clinical development of 17-AAG and the development of geldanamycin analogs. We are collaborating with the NCI to develop analogs of geldanamycin, a polyketide, for the treatment of

6

Table of Contents

cancer. Our lead geldanamycin candidate, 17-AAG, an analog of geldanamycin, is being evaluated in Phase II and Phase Ib clinical trials sponsored by the NCI. We have developed a new formulation of 17-AAG (KOS-953) that we believe addresses a limitation of the first generation formulation. We initiated self-funded Phase I and Phase Ib clinical trials of KOS-953 during 2004. KOS-953 is being evaluated as monotherapy in multiple myeloma and in combination with Velcade^®. Two Phase Ib clinical trials in combination with both Herceptin^® and Gleevec^® are also ongoing. In 2004, 17-AAG received orphan drug status from both the European Medicines Agency (EMEA) and the U.S. Food and Drug Administration (FDA) for the treatment of two hematologic cancers, chronic myelogenous leukemia and multiple myeloma. In addition, we are collaborating with the NCI on developing KOS-1022, a highly-potent, water-soluble and orally-active analog of geldanamycin, for which Phase I clinical trials in advanced malignancies with intravenous administration were initiated in 2004. Under a Kosan-sponsored IND that was filed in December 2004, we plan to initiate Phase I clinical trials of KOS-1022 in hematologic malignancies in the first half of 2005. We plan to submit during 2005 an IND filing to support clinical trials with oral administration of KOS-1022. Under the terms of the CRADAs, we are responsible for the provision of sufficient quantities of clinical materials necessary to conduct the clinical trials funded by the NCI of 17-AAG and other geldanamycin analogs, and we have the option, at our expense, to conduct our own clinical trials of 17-AAG and geldanamycin analogs in addition to those funded by the NCI. The following table summarizes the current development status of our Hsp90 inhibitor program.

Gastrointestinal Motility

One of the actions of erythromycin is stimulation of gastrointestinal movement, or GI motility. Erythromycin-derived compounds, called motilin agonists, may be useful to treat diseases such as gastroparesis and other conditions in which GI motility is impaired or needs to be increased. Some erythromycin analogs have been effective after a single dose but were found to be ineffective after multiple-dose regimens, also known as tachyphylaxis. We have made and tested in preclinical models several series of motilin agonists that stimulate GI motility with minimal tachyphylaxis and from which the unnecessary antibiotic activity of erythromycin has been eliminated. We are currently evaluating a number of these compounds in advanced preclinical studies and plan to enter clinical trials with one of our motilin agonist compounds in 2006.

7

Table of Contents

KOSAN BIOSCIENCES

Ketolide Antibiotics

Macrolides are clinically important antibiotics that have been used for the treatment of many infectious diseases for more than 40 years. Today, an estimated 45 million people in the U.S. have acne, a highly visible bacterial infection of the skin. As the use of existing antibiotics continues, there has been increasing bacterial resistance to these treatments, driving the need in the topical market to develop new and more effective antibiotics. With few novel agents, there exists an opportunity for an effective and safe treatment in the acne market. Ketolides, analogs of the polyketide erythromycin and a class of macrolide antibiotics, not only possess the potency and spectrum of activity shown by current antibiotics, such as erythromycin, clindamycin and tetracycline, but also are effective against some of these resistant organisms. We have generated several proprietary ketolides including our lead ketolide candidate, KOS-1835, and plan to advance the product candidate into clinical trials as a potential topical treatment of skin, ear and eye infections.

Early-Stage Programs

We have three early-stage research programs, based on polyketides. Two are focused on anticancer agents and one is focused on infectious disease. We plan to evaluate these early-stage research programs in 2005 with the intent to advance one or more of these programs into preclinical development in 2006.

Our Technology Platform

Our technology platform has five components: polyketide gene alteration, chemo-biosynthesis, heterologous over-expression, polyketide gene synthesis and polyketide chemical synthesis. Together, our technologies enable us to modify, create and produce proprietary polyketides with potential for development as valuable pharmaceutical products.

Polyketide Gene Alteration

The structure of a polyketide is primarily determined by variation in the number, order and type of modules in the PKS. Our technologies enable us to alter modules in a specific, directed manner, and thus we can make specific and predictable changes to the structure of a polyketide.

Polyketides are structurally complex compounds that are difficult to make or modify chemically. Because each building block of a polyketide is selected by a specific module of the PKS, we use our technologies to make precise structural changes by altering the module that specifies the targeted building block. We use our technologies to improve properties of known, biologically active molecules. This can be done by using our technology to change a portion of the polyketide that cannot be chemically modified to one that can, allowing us to make subsequent chemical modifications not otherwise practicable. We can also make changes in the structure of existing proprietary polyketide products to create a new polyketide proprietary to us.

We also are developing the capability to synthesize polyketide genes to facilitate biosynthetic production and to avoid the need to isolate polyketide synthase genes, or DNA, from nature.

Chemo-Biosynthesis

We incorporate chemically synthesized fragments into complicated polyketide structures, permitting changes in the polyketide structures and properties in ways that have not been achieved by any other process. Microorganisms containing this modified PKS are fed our chemically synthesized fragments that substitute for the natural building-block intermediate.

Heterologous Over-Expression

We can isolate a polyketide gene cluster from one organism and transfer it to another. This is important because many polyketides are produced by microorganisms that are difficult or slow to grow, or in which recombinant DNA

8

Table of Contents

methods have not been developed. In addition, polyketides are often produced in small amounts in organisms that naturally produce them, which can limit their commercial development. We have modified microorganisms, such as S. coelicolor, M. xanthus and E. coli, for efficient polyketide manipulation and production. Our proprietary technologies allow us to transfer polyketide genes to these microorganisms to enable easier manipulation and increased production of polyketides necessary for commercialization.

Polyketide Gene Synthesis

We have developed advanced methods to complete the chemical synthesis of large genes, such as those that encode polyketides. We have used these genes to produce several polyketide synthase genes, and we are using them to create a novel approach toward combinatorial biosynthesis.

Polyketide Chemical Synthesis

We have developed state-of-the-art capabilities to chemically synthesize polyketides and to generate chemical analogs of polyketide structures. This enables us to use total chemical synthesis as an alternative to polyketide gene alteration in order to obtain novel compounds and to combine chemical and biosynthetic approaches for production when appropriate.

Research and Development Expenses

Our research and development activities consist primarily of salaries and other personnel-related expenses, clinical trial-related services performed by clinical research organizations and research institutions and other outside service providers, licensing-related expenses, lab consumables and facility-related expenses. We group these activities into two major categories: “research and preclinical” and “clinical development.” Our research and development expenses were approximately $40.2 million in 2004, $36.8 million in 2003 and $28.4 million in 2002.

Intellectual Property

Our intellectual property consists of patents, copyrights, trade secrets and know-how. Our ability to compete effectively depends in large part on our ability to obtain patents for our technologies and products, maintain trade secrets and operate without infringing the rights of others and prevent others from infringing our proprietary rights. We will be able to protect our technologies from unauthorized use by third parties only to the extent that they are covered by valid and enforceable patents, or are effectively maintained as trade secrets. Accordingly, patents or other proprietary rights are an essential element of our business. As of December 31, 2004, we owned or were exclusively licensed under approximately 97 patents and 139 patent applications in the U.S., approximately eight patents and 39 patent applications in Europe and approximately three patents and 38 patent applications in Japan. In determining these numbers, we do not count European or Japanese patent applications until they enter the national or regional stage from a Patent Cooperation Treaty filing, and when a European patent application issues, we count only one patent regardless of the number of validation countries. We also own or are exclusively licensed under patents and patent applications in other countries and pursue patent protection for each case in countries where we believe it is commercially reasonable and advantageous to do so.

Our policy is to file patent applications to protect technology and compounds commercially important to our business. We also rely on trade secrets to protect our technology where patent protection is deemed inappropriate or unobtainable. We protect our proprietary technology and processes, in part, by confidentiality agreements with our employees, consultants, collaborators and certain contractors. There can be no assurance that our confidential trade secret information will not be disclosed, that others will not independently develop substantially equivalent proprietary information and techniques or that we can meaningfully protect our proprietary information.

9

Table of Contents

KOSAN BIOSCIENCES

Collaborative Research, Development and License Agreements

Roche

In September 2002, we signed a research and collaboration agreement with Roche. Under the terms of the agreement, Roche has worldwide exclusive rights to market and sell KOS-862 and other epothilones in the field of oncology, and we will co-develop and have the right to co-promote the products in the United States. The principal focus of the collaboration is the clinical development and marketing of KOS-862 and a second-generation epothilone, KOS-1584.

Under the agreement, we are entitled to receive clinical development funding for our activities under the collaboration, process and back-up development program funding at specified levels and milestone payments based upon achievement of clinical, regulatory and commercial events. Roche is responsible for the majority of the costs of the epothilone clinical trials conducted under our collaboration and is funding all of the KOS-862 and KOS-1584 clinical trials other than the KOS-862 prostate cancer clinical trial, which we are funding. We estimate that these types of cash payments could total up to $220 million, based upon our current expectations for program expenses and assuming successful development and commercialization of one product that achieves each milestone event. As of December 31, 2004, we had received approximately $63.4 million related to this collaboration. In addition, we are entitled to receive royalties on any sales of collaboration products. We have the option to increase our royalties by contributing to development costs and by co-promotion of products resulting from the collaboration in the United States.

Pursuant to this agreement, we recognized approximately $20.5 million and $26.9 million of contract revenue in 2004 and 2003, respectively. Such amounts, excluding the ratable portion of an initial fee and milestones, approximated research and development expenses under this collaboration agreement. Included in 2004 revenue was $6.0 million in non-recurring milestones achieved under the agreement and the ratable portion of the $25 million initial fee that is being recognized through the second half of 2009, the estimated clinical development period. In December 2004, upon the commencement of a Phase I clinical trial of KOS-1584, our second-generation Epothilone D, we extended the estimated clinical development period from the second half of 2007, which was initially estimated when we entered into the collaboration with Roche in 2002, to the second half of 2009. Our initial estimate of the clinical development period contemplated development of one product candidate, KOS-862.

Either party may terminate the agreement at any time for the other party’s uncured material breach of the agreement. Roche may terminate the agreement as a whole for “cause” relating to the success of the program (not needing to relate to any breach by us). Roche may also terminate the agreement as a whole or in part for convenience. Upon such a termination by Roche for cause or convenience, the licensed rights would revert to us in the terminated territories, and Roche would be required to provide us with certain license rights, data and other assistance to take over the previously licensed products in the terminated territories. Unless earlier terminated, the agreement will remain in effect for the life of the relevant patents or a specified number of years from product launch.

Sloan-Kettering Institute for Cancer Research

Effective August 2000, we signed a research and license agreement with Sloan-Kettering relating to epothilones. Under the agreement, we used our technologies to produce KOS-862 and work collaboratively with Sloan-Kettering to develop compounds and production methods. Under the agreement, each party licensed the other party for activities as part of the collaborative program. In addition, Sloan-Kettering granted us a license for all further development and commercialization of compounds selected by us in accordance with the agreement.

Under the agreement, we paid Sloan-Kettering an initial license fee and funding for the collaborative research program and are required to pay annual maintenance fees and patent fees during the term of the agreement. In addition, we

10

Table of Contents

must pay Sloan-Kettering milestone payments if clinical development milestones are reached, royalty payments based on net sales of products covered by the collaboration, and a share of some sublicensing revenues.

We may terminate the agreement in whole or on a product-by-product basis at any time. Either party may terminate the agreement upon the material breach of the agreement by the other party. Upon termination, all licenses granted by one party to the other revert to the granting party, and we are required to transfer to Sloan-Kettering any marketing authorizations, data and regulatory filings with respect to any compounds for which we do not retain rights under the agreement. Unless earlier terminated, the agreement will remain in effect for the life of the relevant patents or a specified number of years from product launch. Effective September 2002, we received a consent from Sloan-Kettering to enter into a sublicense agreement with Roche and amended the research and license agreement with Sloan-Kettering to harmonize it with the collaboration agreement with Roche.

In 2003, we were in a dispute and litigation with Sloan-Kettering regarding whether certain epothilone analogs developed in Sloan-Kettering laboratories during our collaboration were exclusively licensed to us, the amount of research funding due Sloan-Kettering under the contract and Sloan-Kettering’s right to terminate it. Effective September 2003, we settled the dispute and litigation. Under the terms of the settlement, we and Sloan-Kettering agreed that rights to the disputed epothilone analogs are included in the exclusive license to us under the August 2000 research and license agreement, and that so long as we make the payments required under the settlement, Sloan-Kettering will not have the right to terminate that agreement on the basis of any alleged breach prior to the settlement. The settlement agreement provided for an upfront payment to Sloan-Kettering and additional payments upon the achievement of certain development milestones. In addition, under the settlement, we are obliged to undertake certain development activities with the previously disputed analogs.

Stanford University

Effective March 1996, we entered into an exclusive license agreement with the Board of Trustees of the Leland Stanford Junior University, or Stanford University, for certain technology and related patent rights and materials for the recombinant production of novel polyketides. Under the terms of the agreement, we paid a license issue fee to Stanford University, and we pay annual maintenance fees and patent fees, make milestone payments based on achievement of specified events with licensed products, pay royalties on net sales of products claimed in or originating from the licensed technology and pay a share of some sublicensing fees. In March 2000, we amended the agreement to provide us an exclusive option to acquire an exclusive or non-exclusive license to future patents or patent applications that are related to certain technology developed by one of our founders, Chaitan S. Khosla, Ph.D., related to polyketides or their production, and chosen by Stanford University in its discretion. Under the March 2000 amendment, we paid an issue fee in preferred stock that was subsequently converted to common stock in our initial public offering, and we are obligated to pay an option fee for each patent or patent application included in our option rights. We are also obligated to pay Stanford University a license fee and an annual fee, as well as royalties on net sales and a share of some sublicensing fees under the original agreement, with respect to each patent or patent application for which we exercise our option. Effective September 2002, we received a consent from Stanford University to enter into a sublicense agreement with Roche and amended the research and license agreement to harmonize it with the collaboration agreement with Roche and effect certain other changes.

National Institutes of Health/NCI

Effective October 2002, we entered into two CRADAs with the NIH. Under these agreements, we will collaborate with the NCI in the clinical development of 17-AAG and the development of other geldanamycin analogs, including DMAG. We have the option to obtain exclusive licenses under patents claiming inventions made in the collaborations under standard NIH licensing terms. Under the CRADAs, we must conduct various activities, including providing quantities of the drugs for the preclinical studies and clinical trials, and a specified number of personnel to the

11

Table of Contents

KOSAN BIOSCIENCES

programs. The clinical trials of 17-AAG and other analogs are or will be sponsored by the NCI in collaboration with us or by us solely. Unless earlier terminated, the term of the collaborative programs is currently scheduled for four years in total, but may be adjusted as the programs proceed by agreement with NIH. Either party may unilaterally terminate either or both of the agreements; provided that if we elect to terminate, we will still be required to meet certain obligations to supply licensed product to NIH.

Under a November 2002 license agreement with the NIH, we obtained exclusive commercial rights under patent rights relating to the geldanamycin analogs being developed under the CRADAs. Initially, our license is limited to specified medical fields of use, including treatment of cancer, and the prevention of undesired cell growth or its deleterious effects, such as the prevention of restenosis and neurodegenerative diseases. We intend to include several additional fields of use to the license, subject to NIH’s approval. We have paid the NIH an up-front fee and will owe annual minimum royalties, patent fees, royalties based on net sales of licensed products, milestone payments if clinical development milestones are reached and a portion of our sublicensing revenues. Our license could terminate if we are not diligent in developing licensed products or otherwise default in performing our material obligations under the agreement and fail to cure the deficiency. We may terminate the agreement on a country-by-country basis. The agreement contains standard NIH licensing terms. Unless the agreement is earlier terminated, our license will remain in effect for the life of the licensed patents.

In connection with our collaboration and license agreements for the years ended December 31, 2004, 2003 and 2002, we have expensed approximately $1.0 million, $2.5 million and $1.5 million, respectively, as research and development expenses.

Competition

The pharmaceutical and biotechnology industries are intensely competitive. Many companies, including biotechnology, chemical and pharmaceutical companies, are actively engaged in research and development of drugs for the treatment of the same diseases and conditions as our potential product candidates. Many of these companies have substantially greater financial and other resources, larger research and development staffs and more extensive marketing and manufacturing organizations than we do. In addition, some of them have considerable experience in preclinical testing, clinical trials and other regulatory approval procedures. There are also academic institutions, governmental agencies and other research organizations that are conducting research in areas in which we are working. They may also market commercial products, either on their own or through collaborative efforts.

We face significant competition from large pharmaceutical companies that are pursuing the same or similar technologies, including polyketide manipulation, as the technologies used by us in our drug discovery and production efforts. For example, a number of companies have cloned polyketide synthase genes and claimed or described in patents or publications those genes or technology to modify them or express them in heterologous hosts using recombinant DNA technology. Such companies include, for example, Abbott Laboratories (erythromycin, niddamycin); Dow Agrosciences (spinosyn); Eli Lilly and Company (tylosin, spiramycin); Novartis (soraphen, epothilone); and Merck (avermectin, lovastatin). We also face competition from biotechnology companies and academic researchers that engage in research similar to our own. For example, Biotica Ltd. is a biotechnology company that has published patent applications, issued patents and entered into collaborations with pharmaceutical companies relating to efforts to modify polyketide synthase genes using recombinant DNA technology.

We expect to encounter significant competition for any of the pharmaceutical products we develop. Companies that complete clinical trials, obtain required regulatory approvals and commence commercial sales of their products before their competitors may achieve a significant competitive advantage. We are aware that many other companies or institutions are pursuing the development of drugs and technologies directly targeted at applications for which we are

12

Table of Contents

developing our drug compounds. Competing epothilones in clinical development include those being developed by Bristol-Myers Squibb (reported to be in Phase III clinical trials), Novartis AG (reported to be in Phase II clinical trials) and Schering AG (reported to be in Phase I clinical trials). In Hsp90 inhibitors, Conforma Therapeutics has initiated Phase I clinical trials with its formulation of 17-AAG. In gastrointestinal motility, Chugai Pharmaceuticals has a motilide compound in clinical trials. Thus, it is likely that, even if we are successful in developing any of our product candidates, one or more of these compounds of our competitors will be approved and marketed as products before our own. This could place us and our collaborators at a significant disadvantage, especially if our compounds do not have superior properties or cost advantages, and could prevent us from realizing significant commercial benefit from such products.

Developments by others may render our drug candidates or technologies obsolete or noncompetitive. We face and will continue to face intense competition from other companies for collaborative arrangements with pharmaceutical and biotechnology companies, for establishing relationships with academic and research institutions and for licenses to additional technologies. These competitors, either alone or with their collaborative partners, may succeed in developing technologies or products that are more effective than ours.

To compete successfully we must develop proprietary positions and patented drugs for therapeutic markets that have not been satisfactorily addressed by conventional research strategies and, in the process, expand our technical expertise. We must also successfully pursue our technologies and product candidates outside the scope of proprietary rights of others, or succeed in obtaining licenses under proprietary rights that we are unable to avoid. Our potential products, even if successfully tested and developed, may not be adopted by physicians over other products and may not offer economically feasible alternatives to other therapies.

Government Regulation

The FDA and comparable regulatory agencies in state and local jurisdictions and in foreign countries impose substantial requirements upon the clinical development, manufacture and marketing of pharmaceutical products. These agencies and other federal, state and local entities regulate research and development activities and the testing, manufacture, quality control, safety, effectiveness, labeling, storage, record keeping, approval, advertising and promotion of our potential products.

The process required by the FDA before our products may be marketed in the United States generally involves the following:

The testing and approval process requires substantial time, effort and financial resources, and we cannot be certain that any approvals for any of our potential products will be granted on a timely basis, if at all.

Prior to commencing clinical trials, which are typically conducted in three sequential phases, we must submit an IND application to the FDA. The IND automatically becomes effective 30 days after receipt by the FDA, unless the FDA, within the 30-day time period, raises concerns or questions about the conduct of the trial. In such a case, the IND sponsor and the FDA must resolve any outstanding concerns before the clinical trial can begin. Our submission of an IND may not result in FDA authorization to commence a clinical trial. Further, an independent institutional review

13

Table of Contents

KOSAN BIOSCIENCES

board at the medical center proposing to conduct a clinical trial must review and approve the plan for the clinical trial before it commences.

We may not successfully complete any of the three phases of testing of any of our potential products within any specific time period, if at all. Furthermore, the FDA, an institutional review board or the trial sponsor may suspend a clinical trial at any time on various grounds, including a finding that the subjects or patients are being exposed to an unacceptable health risk.

The results of product development, preclinical studies and clinical trials are submitted to the FDA as part of an NDA or BLA. The FDA may deny an NDA or BLA if the applicable regulatory criteria are not satisfied or may require additional clinical data. Even if such data are submitted, the FDA may ultimately decide that the NDA or BLA does not satisfy the criteria for approval. Once issued, the FDA may withdraw product approval if compliance with regulatory standards is not maintained or if problems occur after the product reaches the market. In addition, the FDA may require testing and surveillance programs to monitor the effect of approved products that have been commercialized, and the FDA has the power to prevent or limit further marketing of a product based on the results of these post-marketing programs.

Satisfaction of FDA requirements or similar requirements of state, local and foreign regulatory agencies typically takes several years, and the actual time required may vary substantially, based upon the type, complexity and novelty of the product or indication. Government regulation may delay or prevent marketing of potential products or for new indications for a considerable period of time and impose costly procedures upon our activities. Success in early-stage clinical trials does not assure success in later-stage clinical trials. Data obtained from clinical activities are not always conclusive and may be susceptible to varying interpretations that could delay, limit or prevent regulatory approval. Even if a product receives regulatory approval, the approval may be significantly limited to specific indications and dosages. Further, even after regulatory approval is obtained, later discovery of previously unknown problems with a product may result in restrictions on the product or even complete withdrawal of the product from the market. Delays in obtaining, or failures to obtain, additional regulatory approvals for any of our products would have a material adverse effect on our business.

Any products manufactured or distributed by us pursuant to FDA approvals are subject to continuing regulation by the FDA, including record-keeping requirements and reporting of adverse experiences with the drug. Drug manufacturers and their subcontractors are required to register their establishments with the FDA and certain state agencies, and are subject to periodic unannounced inspections by the FDA and certain state agencies for compliance with good manufacturing practices, which impose certain procedural and documentation requirements upon us and our third-party manufacturers. We cannot be certain that we or our suppliers will be able to comply with the good manufacturing practices regulations and other FDA regulatory requirements.

Outside the United States, our ability to market a product is contingent upon receiving a marketing authorization from the appropriate regulatory authorities. The requirements governing the conduct of clinical trials, marketing authorization, pricing and reimbursement vary widely from country to country. At present, foreign marketing authorizations are applied for at a national level, although within the European Union, or EU, registration procedures are available to companies wishing to market a product in more than one EU member state. If the regulatory authority is satisfied that adequate evidence of safety, quality and efficacy has been presented, a marketing authorization will be granted. This foreign regulatory approval process involves all of the risks associated with FDA clearance and may be influenced by FDA actions with respect to the product.

14

Table of Contents

Forward-Looking Statements

This annual report on Form 10-K contains forward-looking statements within the meaning of Section 27A of the Securities Act of 1933, as amended, and within the meaning of Section 21E of the Securities Exchange Act of 1934, as amended, which are subject to the “safe harbor” created by those sections. These forward-looking statements include, but are not limited to, statements about:

In some case, forward-looking statements can be identified by words such as “expect,” “anticipate,” “intend,” “believe,” “hope,” “assume,” “estimate,” “plan,” “will” and other similar words and expressions. In addition, any statements that refer to expectations, projections or other characterizations of future events or circumstances are forward-looking statements. Discussions containing these forward-looking statements may be found throughout this annual report on Form 10-K, including the sections entitled “Business” and “Management’s Discussion and Analysis of Financial Condition and Results of Operations.” These forward-looking statements involve risks and uncertainties, including the risks discussed below under the heading “Risks Factors That May Affect Results of Operations and Financial Condition,” that could cause our actual results to differ materially from those in the forward-looking statements. We undertake no obligation to publicly release any revisions to the forward-looking statements or reflect events or circumstances after the date of this document. The risks discussed below under the heading “Risks Factors That May Affect Results of Operations and Financial Condition” and elsewhere in this report should be considered in evaluating our prospects and future financial performance.

Risk Factors That May Affect Results of Operations and Financial Condition

You should carefully consider the risks described below, together with all of the other information included in this report, in considering our business and prospects. The risks and uncertainties described below are not the only ones facing us. Additional

15

Table of Contents

KOSAN BIOSCIENCES

risks and uncertainties not presently known to us or that we currently deem immaterial also may impair our business operations. The occurrence of any of the following risks could harm our business, financial condition or results of operations.

We have a history of net losses and may never become profitable.

We commenced operations in 1996 and are still in an early stage of development. We have not commercialized any products, and we have incurred significant losses to date. As of December 31, 2004, we had an accumulated deficit of approximately $101.2 million. To date, our revenues have been primarily from collaborations and government grant awards. Our expenses have consisted principally of costs incurred in research and development and from general and administrative costs associated with our operations. We have incurred net losses since our inception, including net losses of approximately $22.1 million in 2004, $9.7 million in 2003 and $20.9 million in 2002. We expect our expenses to increase and to continue to incur operating losses for at least the next several years as we continue our research and development efforts for our drug candidates and research programs. The amount of time necessary to successfully commercialize any of our drug candidates successfully is long and uncertain, and successful commercialization may not occur at all. As a result, we may never become profitable.

We expect that additional financing will be required, and an inability to obtain the capital necessary to fund our operations on acceptable terms or at all would threaten the continued operation of our business.

We expect that additional financing will be required in the future to fund operations. We do not know whether additional financing will be available when needed, or that, if available, we will obtain financing on favorable terms. We have consumed substantial amounts of cash to date and expect capital outlays and operating expenditures to increase over the next several years as we expand our infrastructure and research and development activities.

We may raise additional financing through public or private equity offerings, debt financings, additional corporate collaboration or licensing arrangements or any combination of the foregoing or other arrangements. To the extent we raise additional capital by issuing equity securities or other equity-related financing, our stockholders may experience dilution. Debt financing, if available, may subject us to restrictive covenants and significant interest costs. To the extent that we raise additional funds through collaboration and licensing arrangements, it may be necessary to relinquish some rights to our technologies, product candidates or marketing territories. If we are unable to raise sufficient funds when needed, our business, financial condition and results of operations would be materially harmed, and we may not be able to continue developing our products or may need to relinquish rights to product candidates at an earlier stage of development or on less favorable terms than we would otherwise choose. See “Item 7. Management’s Discussion and Analysis of Financial Condition and Result of Operations—Liquidity and Capital Resources.”

We believe that our existing cash and investment securities and anticipated cash flow from our existing collaboration with Roche will be sufficient to support our current operating plan into the second half of 2006, although we may choose to obtain additional financing from time to time. We have based this estimate on assumptions that may prove to be wrong. Our future capital uses and requirements depend on numerous forward-looking factors, including the following:

16

Table of Contents

If our current collaborations are unsuccessful or if conflicts develop with our collaborators, our research and development efforts could be delayed, curtailed or terminated, our revenues could significantly decrease and our operations may be adversely affected.

We have a corporate research and commercialization collaboration with Roche in the field of epothilones. We also have collaborations with, or have licenses to technology and compounds from, several research groups, including Sloan-Kettering in the field of epothilones, the NCI in the field of geldanamycin analogs and Stanford University in the field of polyketide technology. The agreements permit our collaborators or licensors to terminate the agreements under certain circumstances. We may not be able to maintain or extend these collaborations or license agreements on acceptable terms, if at all. If we do not maintain, extend or replace our corporate collaboration with Roche, our research and development efforts could be delayed, our revenues would significantly decrease and our operations could be adversely affected. If we are unable to maintain our research collaborations or if our license agreements are terminated, our research and development efforts could be delayed, curtailed or terminated or we could lose our rights to use the licensed technology and compounds.

We do not control the amount and timing of resources that our collaborators devote to our programs or potential products, nor the scope, content and timing of the efforts that they conduct or permit under the collaborations. As a result, we do not know if our collaborators will dedicate sufficient resources or if the development or commercialization efforts by our corporate partners will be successful. We also do not know if the development or commercialization efforts by our collaborators will be the same as those we would choose to devote if we solely controlled the development and commercialization of our programs and product candidates. In particular, in our collaboration with Roche, we do not control the amount and timing of resources that Roche devotes to the epothilone program beyond limited funding for certain Kosan activities specified under the contract, and we do not control the scope, content and timing of the preclinical studies, clinical trials and other development efforts that Roche conducts or permits under the program. In our collaboration with the NCI, we do not control the selection, conduct, timing and resources provided to clinical trials of geldanamycin analogs sponsored by the NCI. For these reasons, we may choose to undertake product development efforts that are within the scope of our collaborations at our own expense. We also do not know whether our current collaborative partners or future collaborative partners, if any, might pursue alternative technologies or develop alternative products either on their own or in collaboration with others, including our competitors, as a means for developing treatments for the diseases targeted by collaborative arrangements with us. In addition, business combinations or significant changes in a collaborator’s business strategy may adversely affect a collaborator’s willingness or ability to continue the collaboration with us.

Failure by our corporate partners to develop or commercialize a compound or product for which they have rights from us could materially harm our business, financial condition and results of operations. For example, if Roche does not successfully develop and commercialize a product from our epothilone program, we may not receive any future milestone payments and will not receive any royalties under our collaboration with Roche.

17

Table of Contents

KOSAN BIOSCIENCES

If our collaborators fail to conduct the collaborative activities successfully and in a timely manner or if they or our licensors breach or terminate their agreements with us, the development or commercialization of the affected product candidates, technology or research program could be delayed or terminated. If any of our existing collaboration agreements are terminated, we may be required to seek new collaborators or to undertake product development and commercialization at our own expense. This may limit the number of product candidates we will be able to develop and commercialize, significantly increase our capital requirements and reduce the likelihood of successful product introduction. Disputes might also arise with collaborators or licensors concerning rights to particular compounds or technologies. If we are unable to resolve these disputes in our favor, we could lose our rights to use those compounds or technologies.

If we fail to enter into new collaborative agreements in the future, our business and operations would be negatively impacted.

Our strategy depends upon the formation and sustainability of multiple collaborative arrangements and license agreements with third parties. We expect to rely on these arrangements for not only financial resources, but also for expertise that we expect to need in the future relating to clinical trials, manufacturing, sales and marketing, and for license and technology rights. Although we have established collaborative arrangements and various license agreements, we do not know if we will be able to establish additional arrangements on favorable terms, or whether current or any future collaborative arrangements will ultimately be successful. There have been, and may continue to be, a significant number of business combinations among large pharmaceutical companies that have resulted, and may continue to result, in a reduced number of potential future corporate collaborators, which may limit our ability to find partners who will work with us in developing and commercializing our drug candidates. If we do not enter into new collaborative agreements, we may be required to curtail, suspend or terminate research and development programs and therefore our ability to generate revenues from these programs will be adversely affected. Our ability to start new research and development programs may also be materially harmed.

Our potential products are in an early stage of development, and substantial additional effort and expense will be necessary for development.

Our drug candidates are in early stages of research and development. We may not be able to develop products that prove to be safe and effective, meet applicable regulatory standards, are capable of being manufactured at reasonable costs or can be marketed successfully. All of the potential products that we are currently developing will require significant development and investment, including extensive preclinical and clinical testing, before we can submit any application for regulatory approval.

Our products must satisfy rigorous standards of safety and efficacy before they can be approved by the FDA and international regulatory authorities for commercial use. We will need to conduct significant additional research, preclinical testing and clinical trials before we can determine if our products are sufficiently safe and effective to file with the FDA and other regulatory agencies for product approval. Clinical trials are expensive, and therefore, significant amounts of money will need to be spent testing our products.

In addition, significant time and investment will be required to try to develop manufacturing processes for our products so that they are economical to manufacture on a commercial scale and satisfactory to the FDA and other governmental authorities.

The progress and results of our animal and human testing are highly uncertain.

We must provide the FDA and foreign regulatory authorities with clinical data that demonstrate the safety and efficacy of our products before they can be approved for commercial sale. As a result, commercialization of our product candidates depends upon successful completion of preclinical and clinical trials. Preclinical testing and clinical

18

Table of Contents

development are long, expensive and uncertain processes. It may take us a number of years to complete our testing, and failure can occur at any stage of testing. For example, in November 2004, we discontinued a Phase II clinical study of KOS-862 in non-small cell lung cancer because the study did not meet the primary objective of tumor response in the first stage of a two stage clinical trial, and in June 2004, we discontinued a Phase II clinical study of KOS-862 in colorectal cancer due to unanticipated cumulative drug toxicities in patients who had previously been treated with the cancer treatment oxaliplatin. We could experience similar failures in other current or future clinical testing of our product candidates.

Success in preclinical testing and early clinical trials does not ensure that later clinical trials will be successful, and interim results of trials do not necessarily predict final results. A number of companies in the pharmaceutical and biotechnology industry, including Kosan, have suffered significant setbacks in advanced clinical trials, even after promising results in earlier trials. Also, preclinical and clinical data can be interpreted in different ways, which could delay, limit or prevent further testing or regulatory approval.

We do not know whether clinical trials of our product candidates (including ongoing and anticipated clinical trials of KOS-862, KOS-1584, 17-AAG, KOS-953, KOS-1022 or other product candidates) will begin on time or whether any of our clinical trials will be completed on schedule, or at all. Negative or inconclusive results or adverse medical events during a clinical trial could cause a clinical trial to be suspended, repeated or terminated. Certain of the clinical trials of our product candidates are designed to include two stages, with the decision whether to proceed to the second stage dependent on results obtained in the first stage. Failure to achieve predetermined response rates as defined in the protocol may result in the decision not to proceed into the second stage of the related trial.

We have four product candidates in human clinical trials for the treatment of cancer, KOS-862, KOS-1584, KOS-953 and KOS-1022. Anticancer drugs generally have a narrow therapeutic window between efficacy and toxicity. If unacceptable toxicity is observed in a clinical trial, the trial may be terminated at an early stage. For example, in June 2004, we discontinued a Phase II clinical study of KOS-862 in colorectal cancer due to unanticipated cumulative drug toxicities in patients who had previously been treated with the cancer treatment oxaliplatin.

Completion of clinical trials may take several years or more. The length of time generally varies substantially according to the type, complexity, novelty and intended use of the drug candidate. Our clinical trials (including ongoing and anticipated clinical trials of KOS-862, KOS-1584, 17-AAG, KOS-953, KOS-1022 or other product candidates) may be suspended at any time, if we, the FDA, or other regulatory authorities believe the patients participating in our studies are exposed to unacceptable health risks. Our ability to commence or timely complete clinical trials may be adversely affected by many factors, including:

19

Table of Contents

KOSAN BIOSCIENCES

Our product development costs will increase if we have delays in testing or approvals or if we need to perform more or larger clinical trials than planned. If the delays are significant, our financial results and the commercial prospects for our products will be harmed, and our ability to become profitable will be adversely affected. If any clinical trials of our product candidates (including ongoing and anticipated clinical trials of KOS-862, KOS-1584, 17-AAG, KOS-953, KOS-1022 or other product candidates) are not successful, our business, financial condition and results of operations will be harmed.

If we are not able to obtain required regulatory approvals, we will not be able to commercialize our product candidates, and our ability to generate revenue will be materially impaired.

Our product candidates and the activities associated with their development, manufacture and commercialization are subject to comprehensive regulation by the FDA and other regulatory agencies in the United States, and by comparable authorities in other countries. Our products may not be commercialized unless and until our collaborators or we obtain regulatory approval from the FDA or foreign governmental authorities to do so. The process of obtaining regulatory approvals is expensive, often takes many years, if approval is obtained at all, and can vary substantially based upon the type, complexity and novelty of the product candidates involved. We have not received regulatory approval to market any of our product candidates in any jurisdiction and, although our personnel have experience from working at other companies, we as a company have no experience in preparing and filing the applications necessary to gain regulatory approvals to commercialize our products. This lack of experience may impede our ability to obtain FDA or other foreign regulatory approvals to commercialize our products in a timely manner, if at all.

Changes in the regulatory approval policy during the development period, changes in or the enactment of additional regulations or statutes, or changes in regulatory review for each submitted product application, may cause delays in the approval or rejection of an application. Furthermore, the approval procedure and the time required to obtain approval varies among countries and can involve additional testing beyond that required by the FDA. Approval by one regulatory authority does not ensure approval by regulatory authorities in other jurisdictions.

The FDA and other regulatory authorities have substantial discretion in the approval process and may refuse to accept any application or may decide that our data is insufficient for approval and require additional preclinical, clinical or other studies or modifications to the manufacturing processes or facilities or quality control procedures for our products. Any clinical trial may fail to produce results satisfactory to the FDA or other regulatory authorities. Preclinical and clinical data can be interpreted in different ways, which could delay, limit or prevent regulatory approval. Furthermore, even if we file an application with the FDA or other regulatory authorities for marketing approval of a product candidate, it may not result in marketing approval from the FDA or other regulatory authorities.

We do not know whether clinical trials for our product candidates (including ongoing and anticipated clinical trials of KOS-862, KOS-1584, 17-AAG, KOS-953, KOS-1022 or other product candidates) will demonstrate safety and efficacy sufficient to obtain the requisite regulatory approvals or will result in marketable products. Our failure to adequately demonstrate the safety and efficacy of our products under development will prevent receipt of FDA and foreign approvals and, ultimately, commercialization of our products.

20

Table of Contents

We rely on third parties to conduct our clinical trials, and those third parties may not perform satisfactorily.

We do not have the ability to independently conduct clinical trials for our products, and we rely on third parties such as contract research organizations, medical institutions and clinical investigators to perform this function. We also rely on Roche to conduct certain clinical trials for KOS-862 and potentially KOS-1584 and the NCI to conduct certain clinical trials for 17-AAG and KOS-1022. We may rely on future collaborators to conduct clinical trials for our product candidates. If any of these third parties do not successfully carry out their obligations or meet expected deadlines, clinical trials may be extended, delayed or terminated, and our product candidates may not receive regulatory approval or be successfully commercialized.

Even if any of our product candidates receives regulatory approval, we may still face significant development and regulatory difficulties.

Even if the FDA or other regulatory authorities approves a product candidate, the approval may impose significant restrictions on the indicated uses, conditions for use, labeling, advertising, promotion, marketing and/or production of such product, and may impose ongoing requirements for post-approval studies, including additional research and development and clinical trials. In addition, regulatory agencies subject a product, its manufacturer and the manufacturer’s facilities to continual review and periodic inspections. If a regulatory agency discovers previously unknown problems with a product, including adverse events of unanticipated severity or frequency, or problems with the facility where the product is manufactured, a regulatory agency may impose restrictions on that product, our collaborators or us, including requiring withdrawal of the product from the market.

If any of our approved products, our collaborators or we fail to comply with applicable regulatory requirements, a regulatory authority may take various actions, including:

Manufacturing difficulties could delay or preclude commercialization of our products and substantially increase our expenses.

Currently, we are the sole manufacturer of the active pharmaceutical ingredient for KOS-862, we rely on a contract manufacturer for the active pharmaceutical ingredient for KOS-1584 and we use a single outside contractor to formulate drug product for KOS-862 and KOS-1584. We currently maintain limited inventories of the active pharmaceutical ingredient for KOS-862 and KOS-1584 at our facilities in Hayward, California. We currently maintain limited inventories of formulated drug product for KOS-862 and KOS-1584 at our facilities in Hayward, California and at the facilities of an outside contractor. Limited inventories of formulated drug product for KOS-862 are also maintained by Roche.

In our geldanamycin program, we currently use a single manufacturer to make the active pharmaceutical ingredients for 17-AAG (including KOS-953) and KOS-1022. We formulate drug product for 17-AAG, KOS-953 and KOS-1022 both at our own facility and through a contract manufacturer. We currently maintain a limited inventory of 17-AAG,

21

Table of Contents

KOSAN BIOSCIENCES

KOS-953 and KOS-1022 at our facilities in Hayward, California and at the facilities of an outside contractor. The NCI is not obligated to maintain an inventory of either the active pharmaceutical ingredient or formulated drug product for 17-AAG or KOS-1022.

If any of our or our contract manufacturers’ manufacturing or inventory facilities encounter delays, are destroyed or otherwise become unavailable to us, then the clinical development of our product candidates (including KOS-862, KOS-1584, 17-AAG, KOS-953 and KOS-1022) or submissions for their regulatory approval, and therefore commercialization, could be delayed or precluded. Adverse effects would be particularly acute if problems arise with our sole sourcing or inventory relationships. Because our manufacturing processes and those of our contractors are complex and subject to a lengthy regulatory approval process, alternative qualified production capacity may not be available on a timely basis or at all.

A number of factors could cause prolonged interruptions in the manufacturing and supply of our products, including:

While our manufacturing personnel have extensive experience from working at other companies, we as a company have no experience manufacturing products for commercial sale. We may encounter difficulties in attempting to scale-up our manufacturing processes and facilities. We may not be able to achieve such scale-up in a timely manner or at a commercially reasonable cost, if at all. In addition, our facilities in Hayward, California are located within the San Francisco Bay Area, an area that is subject periodically to earthquakes. Our access to any key intermediates, active pharmaceutical ingredient or formulated drug product for our product candidates sourced or inventoried solely through our facilities in Hayward may be subject to interruption in the event of an earthquake.

As discussed above, we rely upon outside contractors to manufacture and supply to us key intermediates, active pharmaceutical ingredients and formulated drug product for our product candidates. Our dependence upon others for the manufacture of our product candidates and components thereof may adversely affect our ability to continue in a timely manner clinical development of our product candidates and may adversely affect any future profit margins and our ability to commercialize any products that we may develop on a timely and competitive basis. Dependence on contract manufacturers involve a number of additional risks, many of which are outside of our control, including:

22

Table of Contents

Any of these factors could cause us to delay or suspend clinical trials, regulatory submissions or commercialization of our products and could result in significantly increased costs.

In addition, our future contract manufacturers may not be in the United States, and we currently utilize a contract manufacturer located outside the United States. Consequently, we may face additional manufacturing difficulties due to a number of potential factors, including importation and customs issues, political uncertainties and a potentially limited ability to enforce our contractual rights against parties not located within the United States.

Any inability to protect our proprietary technologies could significantly harm our business and ability to successfully commercialize product candidates.

Our commercial success will depend in part on our ability to obtain patents and maintain adequate protection of other intellectual property for our technologies and products in the United States and other countries and prevent others from infringing our proprietary rights. If we do not adequately protect our intellectual property, competitors may be able to use our technologies and erode or negate any competitive advantage we may have. The laws of some foreign countries do not protect our proprietary rights to the same extent as the laws of the United States, and we may encounter significant problems in protecting our proprietary rights in these countries.

We will be able to protect our proprietary rights from unauthorized use by third parties only to the extent that our proprietary technologies and products are covered by valid and enforceable patents or are effectively maintained as trade secrets. However, the patent positions of biotechnology companies, including our patent position, involve complex legal and factual questions, and, therefore, we cannot predict with certainty whether our patent applications will be approved or any resulting patents will be enforced. Patents may be challenged, deemed unenforceable, invalidated or circumvented. Thus, any patents that we own or license from third parties may not provide protection against competitors.

The degree of future protection for our proprietary rights is uncertain, and we cannot ensure that:

We apply for patents covering our technologies and drug candidates, as we deem appropriate. However, we may fail to apply for patents on important technologies or products in a timely fashion or at all. Our existing patents and any

23

Table of Contents

KOSAN BIOSCIENCES

future patents we obtain may not be sufficiently broad to prevent others from practicing our technologies or from developing competing products and technologies. For example, our lead product in our geldanamycin analog program, 17-AAG, is not covered by a composition of matter patent, and thus others could develop products containing 17-AAG. We are aware of at least one other company that has been developing product candidates containing 17-AAG. Other competitors may be currently developing, or may in the future develop, products containing 17-AAG. In addition, we generally do not control the patent prosecution of technology that we license from others. Accordingly, we are unable to exercise the same degree of control over this intellectual property as we would over our own.

In addition to patents, we rely on trade secrets and proprietary know-how. We have taken measures to protect our confidential information and trade secrets. However, these measures may not provide adequate protection. We seek to protect our confidential information and trade secrets by entering into confidentiality agreements with employees, collaborators, consultants and others. Nevertheless, parties may breach these agreements or competitors may otherwise obtain or independently develop our trade secrets.

Claims by third-parties of intellectual property infringement would require us to spend time and money and could deprive us of valuable rights needed to develop or commercialize our products.

Our commercial success depends significantly on not infringing the patents and proprietary rights of other parties and not breaching any licenses that we have entered into with regard to our technologies and products. Other parties may currently or in the future possess intellectual property rights covering drug candidates that we are developing or desire to develop and genes, gene fragments, compounds and technologies we use or may wish to use. Any infringement of patent rights or violation of other proprietary rights may require us to obtain a license from another party, forego product development or commercialization or face lawsuits or other claims.

The biotechnology industry is characterized by extensive litigation regarding patents and other intellectual property rights. We are aware of patents and published patent applications that, if valid, and if we are unsuccessful in circumventing or acquiring the rights to these patents, may block our ability to commercialize products based on the drug candidates that we are developing or pursue our PKS (polyketide synthase) gene manipulation and production technologies. We cannot be sure that other parties have not filed for or been issued relevant patents that could affect our ability to obtain patents or successfully operate our business. Others may challenge our patent or other intellectual property rights or sue us for patent infringement, misappropriation of their intellectual property rights or breach of license agreements. We may be required to commence legal proceedings to resolve our patent or other intellectual property rights. An adverse determination in any litigation or in administrative proceeding to which we may become a party could subject us to significant liabilities, result in our patents being deemed invalid, unenforceable or revoked, require us to license disputed rights from others or to cease using the disputed technology. In addition, our involvement in any of these proceedings may cause us to incur substantial costs and result in diversion of management and technical personnel.

We are aware of a significant number of patents and patent applications relating to aspects of our technologies and compounds filed by, and issued to, other parties. Others have filed patent applications or have been granted patents claiming inventions also claimed or licensed by us, and we may have to participate in an interference or other proceeding before a patent agency or court to determine priority of invention or which party was first to invent and, thus, the right to a patent for these inventions. For example, we believe one or more interferences may be declared between patents and applications we own or have exclusively licensed and patents and applications owned by Novartis relating to epothilone biosynthetic genes and Epothilone D; patent applications believed by us to be licensed to Bristol-Myers Squibb relating to epothilones; and patents and applications owned by Abbott Laboratories and Biotica Ltd. relating to erythromycin polyketide synthase genes, methods for altering polyketide synthase genes and

24

Table of Contents

erythromycin analogs. Biotica Ltd. has filed an opposition against one of our European patents related to the recombinant production of polyketides; we have filed a response and a hearing has been scheduled at the European Patent Office in the first half of 2005. In addition, the European Patent Office has recently granted European patents to the German Research Centre for Biotechnology (GBF), which we believe to be licensed to Bristol-Myers Squibb and, if valid in individual European countries, would cover Epothilone D in those countries. Formal oppositions to these patents have been filed with the European Patent Office and we expect a response by GBF to those oppositions beginning the second half of 2005. We are also aware that the United States Patent Office has recently issued a Notice of Allowance to a patent application owned by Oregon State University, which relates to KOS-1584. A proceeding or a lawsuit in which we are alleged to have infringed an issued patent could result in substantial cost to us even if the outcome is favorable, and if the outcome is unfavorable, we could be required to license the other party’s rights, on terms that may be unfavorable to us, or cease using the technology. Even if successful on priority grounds, an interference may result in loss of claims based on patentability grounds raised in the interference. Although patent and intellectual property disputes in the biotechnology area are often settled through licensing or similar arrangements, costs associated with these arrangements may be substantial and could include ongoing royalties. Furthermore, we cannot be certain that a license would be available to us on satisfactory terms, if at all. Companies and others developing products that could compete with our product candidates, such as Bristol-Myers Squibb and Novartis in the area of potential epothilone products, may be particularly unwilling to grant us a license at any price.

Other parties may obtain patents in the future and claim that the use of our technologies infringes these patents or that we are employing their proprietary technology without authorization. We could incur substantial costs and diversion of management and technical personnel in defending ourselves against any claims that the use of our technologies infringes any patents, defending ourselves against any claim that we are employing any proprietary technology without authorization or enforcing our patents against others. Furthermore, parties making claims against us may be able to obtain injunctive or other equitable relief that could effectively block our ability to develop, commercialize and sell products, and could result in the award of substantial damages against us. In the event of a successful claim of infringement against us, we may be required to:

We may not be able to obtain licenses from other parties at a reasonable cost, or at all. If we are not able to obtain necessary licenses at a reasonable cost or at all, we could encounter substantial delays in product introductions while we attempt to develop alternative methods and products, which we may not be able to accomplish.

Litigation or the failure to obtain licenses could prevent us from manufacturing or commercializing products and could materially harm our business, financial condition and results of operation.

If we are unable to recruit and retain skilled employees and consultants, we may not be able to successfully operate our business.

Retaining our current employees and recruiting qualified scientific personnel to perform future research and development work, as well as manufacturing and key management personnel, will be critical to our success. We may also need to hire personnel with expertise in clinical testing, government regulation, marketing, law and finance. Competition is intense among biotechnology, pharmaceutical and healthcare companies, universities and non-profit research institutions for experienced scientists and other personnel, and we may not be able to retain or recruit

25

Table of Contents

KOSAN BIOSCIENCES

sufficient skilled personnel on acceptable terms to allow us to pursue collaborations and develop our products and core technologies to the extent otherwise possible.

In December 2004, the Financial Accounting Standards Board, or FASB, issued Statement No. 123 (revised 2004) “Share-Based Payment,” or SFAS 123R, which will require us to record a charge to earnings for the compensation cost relating to share-based payment transactions, including stock options and employee stock purchase plans, effective July 1, 2005. We are currently evaluating option valuation methodologies and assumptions permitted by the FASB for purposes of implementing the change in accounting treatment. This change will have a negative impact on our financial results. For an illustration of the effect of such a change in our recent financial results, see Note 1 to the Notes to Financial Statements—Summary of Significant Accounting Policies: Stock Options Plans. In addition, new regulations implemented by The Nasdaq National Market requiring shareholder approval for all stock option plans, as well as new regulations implemented by the New York Stock Exchange prohibiting NYSE member organizations from giving a proxy to vote on equity-compensation plans unless the beneficial owner of the shares has given voting instructions, could make it more difficult for us to grant options to employees in the future. To the extent that new regulations make it more difficult or expensive to grant stock options to employees, we may incur increased compensation costs, change our equity compensation strategy or find it difficult to attract, retain and motivate employees, each of which could materially and adversely affect our business.

We face intense competition from large pharmaceutical companies, biotechnology companies and academic groups.

We face, and will continue to face, intense competition from organizations such as large biotechnology and pharmaceutical companies, as well as academic and research institutions and government agencies, that are pursuing competing technologies and products. These organizations may develop or currently possess technologies or products that are superior alternatives to ours. For example, competing epothilones in clinical development include those being developed by Bristol-Myers Squibb (reported to be in Phase III clinical trials), Novartis AG (reported to be in Phase II clinical trials) and Schering AG (reported to be in Phase I clinical trials). In Hsp90 inhibitors, Conforma Therapeutics has initiated Phase I clinical trials with their formulation of 17-AAG. Further, our competitors in the polyketide gene-engineering field may be more effective at implementing their technologies to develop commercial products or may hold or develop patents or other proprietary rights that may prevent us from practicing our technologies and pursuing our programs. Some of these competitors have entered into collaborations with leading companies within our target markets to produce polyketides for commercial purposes.

Any products that we develop through our technologies will compete in multiple, highly competitive markets. Development of pharmaceutical products requires significant investment and resources. Many of the organizations competing with us in the markets for such products have greater capital resources, research and development and marketing staffs, facilities and capabilities, and greater experience in discovery and developing drugs, obtaining regulatory approvals and product manufacturing and marketing. Accordingly, our competitors may succeed in more rapidly developing and marketing technologies and products that are more effective than our technologies and products or that would render our products or technologies obsolete or noncompetitive.

We believe that our ability to successfully compete will depend on, among other things:

26

Table of Contents

If we face product liability claims, these claims will divert our management’s time and we will incur litigation costs, and if we are held liable, our business, financial condition and results of operation may be materially harmed.

We face an inherent business risk of liability claims in the event that the use of our potential products in clinical trials or otherwise, or any other products manufactured in our facility, results in personal injury or death. Even though we have obtained product liability insurance, it may not be sufficient to cover claims that may be made against us. Product liability insurance is expensive, difficult to obtain and may not be available in the future on acceptable terms, if at all. Any claims against us, regardless of their merit, could materially and adversely affect our business, financial condition and results of operation, because litigation related to these claims would strain our financial resources in addition to consuming the time and attention of our management. If we are sued for any injuries caused by our products or products manufactured at our facility, our liability could exceed our total assets.

We use hazardous chemicals and radioactive and biological materials in our business. Any claims relating to improper handling, storage or disposal of these materials could be time consuming and costly.

Our research and development processes involve the controlled use of hazardous materials, including hazardous chemicals and radioactive and biological materials. Some of these materials may be novel, including bacteria with novel properties and bacteria that produce biologically active compounds. Our operations also produce hazardous waste products. We cannot eliminate the risk of accidental contamination or discharge and any resultant injury from these materials. Federal, state and local laws and regulations govern the use, manufacture, storage, handling and disposal of these materials. We could be subject to civil damages in the event of an improper or unauthorized release of, or exposure of individuals to, hazardous materials. In addition, we could be sued for injury or contamination that results from our use or the use by third parties or our collaborators of these materials, and our liability may exceed our total assets. Compliance with environmental laws and regulations may be expensive, and current or future environmental regulations may impair our research, development or commercialization efforts.

We have a stockholders rights plan and anti-takeover provisions in our corporate charter documents that may result in outcomes with which you do not agree.

Our board of directors has the authority to issue up to 10,000,000 shares of preferred stock and to determine the rights, preferences, privileges and restrictions of those shares without further vote or action by our stockholders. The rights of the holders of any preferred stock that may be issued in the future may adversely affect the rights of the holders of common stock. The issuance of preferred stock could make it more difficult for third parties to acquire a majority of our outstanding voting stock.

Our certificate of incorporation provides for staggered terms for the members of the board of directors and prevents our stockholders from acting by written consent. These provisions and other provisions of our bylaws and of Delaware law applicable to us could delay or make more difficult a merger, tender offer or proxy contest involving us. This could reduce the price that investors might be willing to pay for shares of our common stock and result in the market price being lower than it would be without these provisions. In addition, these provisions may frustrate or prevent any attempts by our stockholders to replace or remove our current management by making it more difficult for stockholders to replace members of our board of directors. This is because our board of directors is responsible for appointing the members of our management team.

27

Table of Contents

KOSAN BIOSCIENCES

We have adopted a rights agreement under which all stockholders have the right to purchase shares of a new series of preferred stock at an exercise price of $70.00 per one one-hundredth of a share, if a person acquires more than 20% of our common stock. The rights plan could make it more difficult for a person to acquire a majority of our outstanding voting stock. The rights plan could also reduce the price that investors might be willing to pay for shares of our common stock and result in the market price being lower than it would be without the rights plan. In addition, the existence of the rights plan itself may deter a potential acquiror from acquiring us. As a result, either by operation of the rights plan or by its potential deterrent effect, mergers and acquisitions of us that our stockholders may consider in their best interests may not occur.

Some of our existing stockholders can exert control over us and may not make decisions that are in the best interest of all stockholders.

Our officers, directors and their affiliates together controlled approximately 26% of our outstanding common stock as of December 31, 2004. As a result, these stockholders, if they act together, are able to exert a significant degree of influence over our management and affairs and over matters requiring stockholder approval, including the election of directors and approval of significant corporate transactions. This concentration of ownership may delay or prevent a change in control of us and might affect the market price of our common stock, even when a change may be in the best interests of all stockholders. In addition, the interests of this concentration of ownership may not always coincide with our interests or the interests of other stockholders, and accordingly, they could cause us to enter into transactions or agreements, which we would not otherwise consider.

Our stock price has been, and may continue to be, extremely volatile.

The trading price of our common stock has been, and is likely to continue to be, highly volatile. During the period from December 31, 2003 through December 31, 2004, our common stock traded between $5.71 and $14.77 on the Nasdaq National Market. The trading price of our common stock could be subject to wide fluctuations in price in response to various factors, many of which are beyond our control, including:

28

Table of Contents

In addition, the stock market in general, and the Nasdaq National Market and the market for biotechnology companies in particular, has experienced extreme price and volume fluctuations that have often been unrelated or disproportionate to the operating performance of those companies. These broad market and industry factors may seriously harm the market price of our common stock, regardless of our operating performance. In the past, following periods of volatility in the market price of a company’s securities, securities class-action litigation has often been instituted against that company. If this type of litigation were instituted against us, we would be faced with substantial costs and management’s attention and resources would be diverted, which could in turn seriously harm our business, financial condition and results of operations.

We expect that our quarterly results of operations will fluctuate, and this fluctuation could cause our stock price to decline, creating investor losses.

Our quarterly operating results have fluctuated in the past and are likely to do so in the future. These fluctuations could cause our stock price to fluctuate significantly or decline. Some of the factors that could cause our operating results to fluctuate include:

We expect a large portion of our expenses to be relatively fixed, including expenses for facilities, equipment and personnel. Accordingly, if revenues decline or do not grow due to expiration, termination or amendment of current or future collaboration agreements, licenses or government research grants, failure to obtain new contracts or other factors, we may not be able to reduce our operating expenses correspondingly. In addition, we expect operating expenses to continue to increase. Failure to achieve anticipated levels of revenues could therefore significantly harm our operating results for a particular fiscal period.

Due to the possibility of fluctuations in our revenues and expenses, we believe that quarter-to-quarter comparisons of our operating results are not a good indication of our future performance. Our operating results in some quarters may not meet the expectations of stock market analysts and investors. In that case, our stock price would probably decline.

Changes in the accounting treatment of stock options will adversely affect our results of operations.

Changes in the accounting treatment of stock options will require us to account for employee stock options as compensation expense on our financial statements. In December 2004, the Financial Accounting Standards Board, or FASB, issued Statement No. 123 (revised 2004) “Share-Based Payment,” or SFAS 123R, which will require us to record a charge to earnings for the compensation cost relating to share-based payment transactions, including stock options and employee stock purchase plans, effective July 1, 2005. We are currently evaluating option valuation methodologies and assumptions permitted by the FASB for purposes of implementing the change in accounting treatment. This change will have a negative impact on our financial results. For an illustration of the effect of such a change in our recent financial results, see Note 1 to the Notes to Financial Statements—Summary of Significant Accounting Policies: Stock Options Plans.

29

Table of Contents

KOSAN BIOSCIENCES

If we are unable to favorably assess the effectiveness of internal controls over financial reporting, or if our independent auditors are unable to provide an unqualified attestation report on our assessment, our stock price could be adversely affected.

Pursuant to Section 404 of the Sarbanes-Oxley Act of 2002, on an annual basis, our management is required to report on, and our independent auditors to attest to, the effectiveness of our internal controls over financial reporting. The rules governing the standards that must be met for management to make its annual assessment are complex and require significant documentation and testing. While our internal controls over financial reporting were deemed effective by both our management and our independent auditors as of December 31, 2004, there may be changes in our systems, processes or operations that will effect the effectiveness of internal controls in the future. Because of its inherent limitations, internal control over financial reporting may not prevent or detect misstatements. Our future assessments of internal controls may continue to result in increased expenses and the devotion of significant management resources. If we cannot favorably assess the effectiveness of our internal controls over financial reporting in the future, or if our independent auditors are unable to provide an unqualified attestation report on our assessment, investor confidence and our stock price could be adversely affected.

Employees

As of February 28, 2005, we had 136 full-time employees, 55 of whom hold Ph.D. degrees and 108 of whom were engaged in research and development activities. None of our employees is represented by a collective bargaining agreement, nor have we experienced any work stoppage. We believe that our relations with our employees are good.

Directors and Executive and Other Officers of the Registrant

Our directors and executive and other officers, and their ages as of March 15, 2005, are as follows:

Daniel V. Santi, M.D., Ph.D., is one of our co-founders and has served as Chairman of the Board of Directors since our inception in January 1995 and as our Chief Executive Officer since November 1998. Until his retirement in January 2001, Dr. Santi was on leave of absence from his position as Professor of Biochemistry and Biophysics, and of Pharmaceutical Chemistry at University of California, San Francisco, a position that he held since 1974. Dr. Santi was one of the original members of the Scientific Advisory Boards of Chiron Corporation and Mitotix, Inc., and has served as a consultant to several large pharmaceutical companies. In 1988, Dr. Santi founded and served as Chairman of the Board of Directors of Protos, Inc., a biotechnology firm and a subsidiary of Chiron Corporation, which was merged

30

Table of Contents

with Chiron in 1992. Dr. Santi was also founder and Chairman of Parnassus Pharmaceuticals. Dr. Santi has published over 275 scientific papers and is inventor on many patents in combinatorial chemistry and other areas. Dr. Santi received a B.S. in pharmacy from the State University of New York, an M.D. from the University of California, San Francisco and a Ph.D. in medicinal chemistry from the State University of New York.

Robert G. Johnson, Jr., M.D., Ph.D., has served as our Executive Vice President, Development since April 2004 and as our Chief Medical Officer since January 2003. From January 2002 to April 2004, Dr. Johnson served as Senior Vice President, Medical Affairs and Corporate Development and from September 2000 to January 2002, Dr. Johnson served as Vice President, Medical Affairs and Corporate Development. From 1998 to September 2000, Dr. Johnson was employed by Chiron Corporation, where he served as Vice President, Pharmacology and Preclinical Affairs through 1999 and most recently as Vice President, Corporate Development. From 1991 to 1998, Dr. Johnson was Director of Pharmacology at Merck & Co., Inc., a pharmaceutical company. In addition, Dr. Johnson was a member of the faculty at the University of Pennsylvania from 1987 to 1991 and at Harvard Medical School from 1985 to 1987. Dr. Johnson received his B.A. and Ph.D. in biophysics and his M.D. from the University of Pennsylvania.

Susan M. Kanaya has served as our Senior Vice President, Finance and Chief Financial Officer since January 2002 and was our Secretary from June 2001 to June 2004. From November 1999 to January 2002, Ms. Kanaya served as our Vice President, Finance and Chief Financial Officer. From 1994 to November 1999, Ms. Kanaya was employed by SUGEN, Inc., a publicly held biotechnology company that was acquired by Pharmacia in 1999, most recently serving as Vice President, Finance and Treasurer. Before joining SUGEN, Ms. Kanaya was the Controller at 50/50 Micro Electronics, Inc., an electronics company, and at Power Up Software Corporation, a computer software company. Ms. Kanaya received a B.S. in business administration from the University of California, Berkeley.

Bruce E. MacMillan has served as our Senior Vice President, General Counsel and Secretary since June 2004. From 2003 to June 2004, Mr. MacMillan was an independent consultant. From 2002 to 2003, Mr. MacMillan was Vice President and General Counsel at Caliper Technologies. From 2000 to 2002, Mr. MacMillan was an independent consultant to technology companies. From April 1999 to 2000, Mr. MacMillan was Vice President and General Counsel at SUGEN, Inc. From January 1991 to April 1999 Mr. MacMillan served as Associate General Counsel for Raychem Corporation, an electronics and telecommunications component business. Mr. MacMillan was also General Counsel for Ericsson-Raynet, a joint venture between Raychem and Telefonaktiebolaget L.M. Ericsson. Prior to that he was in private practice for 10 years. Mr. MacMillan holds a B.A. in Economics and an M.B.A. from University of California, Berkeley, and a J.D. from University of California, Hastings College of Law.

Pieter Timmermans, Ph.D., has served as our Senior Vice President, Preclinical Development since January 2005. From August 2004 to January 2005, Dr. Timmermans served as Vice President, Pharmacology and Preclinical Development at Amgen Inc. subsequent to its acquisition of Tularik Inc. From 1997 to 2004, Dr. Timmermans held the same position at Tularik. Previously, he served in various management positions at The DuPont Merck Pharmaceutical Company and E.I. du Pont de Nemours & Company. He also was Associate Professor in Pharmacology at the University of Amsterdam. The author of more than 580 scientific publications, Dr. Timmermans holds a Ph.D. in molecular pharmacology and a B.S. in chemistry, mathematics, physics and biology from Leiden State University in the Netherlands.

Bruce A. Chabner, M.D., has served as a director since September 2001. Dr. Chabner has served as the Chief of Hematology/Oncology at the Massachusetts General Hospital and as Professor of Medicine at Harvard Medical School since 1995. Dr. Chabner has also served as the Associate Director for Clinical Science of the Dana-Farber/Harvard Cancer Center since 1999 and has held numerous academic appointments, including the position of Director of the Division of Cancer Treatment of the National Cancer Institute from 1982 to 1995. Dr. Chabner has received

31

Table of Contents

KOSAN BIOSCIENCES

numerous awards, including Phi Beta Kappa, Alpha Omega Alpha, the Public Health Service’s Distinguished Service Medal, the Karnofsky Award of the American Society for Clinical Oncology and the Bruce F. Cain Award for Drug Development of the American Association for Cancer Research. Dr. Chabner received a B.A. from Yale College and an M.D. from Harvard Medical School.

Peter Davis, Ph.D., has served as a director since April 1998. Since 2002, Dr. Davis has worked as an independent consultant to a number of companies. Dr. Davis served as president of DNA Plant Technologies Corp., an agriculture biotechnology company, from 2001 to 2002. Dr. Davis was a member of the Executive Committee of Pulsar International, S.A., a management consultant company and an affiliate of A.G. Biotech Capital, from 1993 to 2001. From 1975 to 1993, Dr. Davis was a faculty and staff member of the Wharton School of the University of Pennsylvania. His primary appointments included Director of the Applied Research Center and Director of Executive Education. He is a member of the board of directors of several private companies. Dr. Davis received a B.A. in physics from Cambridge University, a Masters Degree in operations research from the London School of Economics and a Ph.D. in operations research from the Wharton School.

Jean Deleage, Ph.D., has served as a director since April 1996. Dr. Deleage is a founder and managing director of Alta Partners, a venture capital firm investing in information technologies and life science companies. Alta Partners was formed in 1996. In 1979, Dr. Deleage founded, and was a managing partner of Burr, Egan, Deleage & Co., a major venture capital firm in San Francisco and Boston. Dr. Deleage was a member of Sofinnova’s initial team, a venture capital organization in Paris, and in 1976 formed Sofinnova, Inc. (the U.S. subsidiary of Sofinnova). Dr. Deleage is presently a member of the board of directors of Rigel Pharmaceuticals, Inc., Xcyte Therapies, Inc. and several private companies. In 1984, Dr. Deleage was awarded the Ordre National du Merite, and in 1993, he was awarded the Legion of Honor from the French government in recognition of his career accomplishments. Dr. Deleage received a Master’s Degree in electrical engineering from Ecole Superieurie d’Electricite and a Ph.D. in economics from the University of Paris, Sorbonne.

Charles J. Homcy, M.D., has served as a director since April 2003. Since November 2003, Dr. Homcy has served as Chief Executive Officer of Portola Pharmaceuticals, Inc., a biopharmaceutical company. From January 2003 to November 2003, Dr. Homcy served as Senior Research and Development Advisor of Millennium Pharmaceuticals. From February 2002 to December 2002, Dr. Homcy served as the President of Research and Development at Millennium Pharmaceuticals. From 1995 to February 2002, he served as Executive Vice President, Research and Development of COR Therapeutics, Inc., where he served as a member of the board of directors from 1998 to February 2002. From 1994 to March 1995, Dr. Homcy was President of the Medical Research Division of American Cyanamid Company-Lederle Laboratories (now a division of Wyeth-Ayerst Laboratories). From 1990 to 1994, Dr. Homcy was Executive Director of the Cardiovascular and Central Nervous System Research Section at Lederle Laboratories. Dr. Homcy currently serves on the boards of directors of Millennium Pharmaceuticals and Cytokinetics, Incorporated, a biopharmaceutical company. Dr. Homcy received his B.A. and his M.D degrees from the Johns Hopkins University in Baltimore

Chaitan S. Khosla, Ph.D., is one of our co-founders and has served as a director since our inception in January 1995. Dr. Khosla has been a Professor of Chemical Engineering, Chemistry and Biochemistry at Stanford University since 2001 and has been a faculty member since 1992. Dr. Khosla is the chairman of our Scientific Advisory Board. Dr. Khosla is the inventor of the combinatorial biosynthesis technology that we licensed from Stanford University. He is the recipient of several awards, including the 1999 Alan T. Waterman award by the National Science Foundation, the 1999 Eli Lilly Award in biological chemistry and the 2000 ACS Award in pure chemistry. Dr. Khosla is the author of over 150 publications and is an inventor on numerous patents. Dr. Khosla received a B.S. Tech. from the Indian Institute of Technology, Bombay, India and a Ph.D. from the California Institute of Technology.

32

Table of Contents

Christopher T. Walsh, Ph.D., has served as a director since April 1996. Dr. Walsh has been the Hamilton Kuhn Professor of Biological Chemistry and Molecular Pharmacology at Harvard Medical School since 1991 and formerly was President of the Dana-Farber Cancer Institute and Chairman of the Department of Biological Chemistry and Molecular Pharmacology at Harvard Medical School. He has performed extensive research in enzyme stereochemistry, reaction mechanisms and the mechanisms of action of anti-infective and immunosuppressive agents. From 1996 until July 2003, Dr. Walsh served as co-chairman of our Scientific Advisory Board. Dr. Walsh is also a member of the board of directors of Vicuron Inc. and several private companies. Dr. Walsh received an A.B. in biology from Harvard University and a Ph.D. in life sciences from The Rockefeller University, New York.

Scientific Advisory Board

The following individuals are members of our Scientific Advisory Board, or SAB:

Chaitan S. Khosla, Ph.D., is the chairman of our SAB and a member of our board of directors.

Paul S. Anderson, Ph.D., has extensive experience in the field of drug discovery and development. Dr. Andersen recently retired from Bristol-Myers Squibb, where he held the position of Vice President, Drug Discovery.

Homer A. Boushey, M.D., is a Professor of Medicine at the University of California, San Francisco. Dr. Boushey is an expert in clinical research on the causes and treatment of asthma and serves as Principal Investigator for University of California, San Francisco’s Asthma Clinical Research Center.

David E. Cane, Ph.D., is Professor of Chemistry and Biochemistry at Brown University. Dr. Cane is an expert in the biosynthesis of natural products, with particular emphasis on macrolide polyketides and terpenes.

Samuel J. Danishefsky, Ph.D., is Professor of Chemistry at Columbia University. Dr. Danishefsky is an expert in synthetic organic chemistry.

Sir David A. Hopwood, Ph.D., is Professor and Head of the Genetics Department at John Innes Institute, Norwich, U.K. Dr. Hopwood is an expert in streptomyces genetics, molecular biology and the genetic manipulation of polyketide genes.

Jack F. Kirsch, Ph.D., is a Professor of Chemistry, Biochemistry and Molecular Biology at the University of California, Berkeley. Dr. Kirsch is an expert in the physical organic chemistry of proteins and enzyme kinetics.

Ivan M. Kompis, Ph.D., has an extensive background in natural products chemistry, in particular antibacterial agents. Dr. Kompis recently retired from Hoffmann-La Roche, where he held the position of Deputy Director of the Department of Infectious Diseases since 1987.

Mohamed A. Marahiel, Ph.D., is Professor of Biochemistry at Philipps University, Marburg, Germany. Dr. Marahiel is an expert in the field of non-ribosomal peptide biosynthesis.

Theo L. Peeters, Ph.D., is Professor at the University of Leuven, Belgium, and chairman of the Belgian Gastroenterology Research Group. He is an expert in the regulation of the contractility of gastrointestinal system.

Edward A. Sausville, M.D., Ph.D., is Associate Director for Clinical Research of the Greenbaum Cancer Center of the University of Maryland, Professor of Medicine and Pharmacology in the University of Maryland School of Medicine, and Adjunct Professor in the University of Maryland School of Pharmacy. Dr. Sausville’s areas of expertise include cancer biology, and cancer drug development and therapy.

Haruo Seto, Ph.D., is Professor Emeritus, University of Tokyo, Japan. Dr. Seto has an extensive background in the structure, biosynthesis and screening of antibiotics.

33

Table of Contents

KOSAN BIOSCIENCES

Amos B. Smith, III, Ph.D., is the Rhodes-Thompson Professor of Chemistry at the University of Pennsylvania. Dr. Smith is an expert in natural products synthesis, bioorganic chemistry and materials science.

Jon S. Thorson, Ph.D., is Professor of Pharmaceutical Sciences at the University of Wisconsin, Madison. He is an expert in the biosynthesis of natural products, with emphasis on pluramycins, enediynes, indolocarbazoles and natural product glycosylation.

Available Information

We maintain a website at www.kosan.com; however, information found on, or that can be accessed through, our website is not incorporated by reference into this report. We make available free of charge on or through our website our annual report on Form 10-K, quarterly reports on Form 10-Q, current reports on Form 8-K and amendments to those reports filed or furnished pursuant to Section 13(a) or 15(d) of the Exchange Act as soon as reasonably practicable after we electronically file such material with, or furnish it to, the SEC. Further, a copy of this report is located at the SEC’s Public Reference Room at 450 Fifth Street, NW, Washington, D.C. 20549. Information on the operation of the Public Reference Room can be obtained by calling the SEC at 1-800-SEC-0330. The SEC maintains a website that contains reports, proxy and information statements, and other information regarding our filings, at www.sec.gov.

In 2002, we adopted a code of ethics that applies to our employees, officers and directors and in February 2004, we amended and restated the code. We have posted the text of our code of ethics on our website at www.kosan.com in connection with “Investor Relations” materials. In addition, we intend to promptly disclose (1) the nature of any amendment to our code of ethics that applies to our principal executive officer, principal financial officer, principal accounting officer or controller, or persons performing similar functions and (2) the nature of any waiver, including an implicit waiver, from a provision of our code of ethics that is granted to one of these specified officers, the name of such person who is granted the waiver and the date of the waiver on our website in the future.

ITEM 2. PROPERTIES

Our facilities consist of approximately 118,000 square feet of research and office space located in Hayward, California, of which approximately 44,000, 69,000 and 5,000 square feet are leased to us until 2013, 2008 and 2005, respectively. We have an option to renew our lease on the 44,000 square foot facility for one additional period of five years and an option to renew our lease on the 69,000 square foot facility for two additional periods of five years.

ITEM 3. LEGAL PROCEEDINGS

None.

ITEM 4. SUBMISSION OF MATTERS TO A VOTE OF SECURITY HOLDERS

None.

34

Table of Contents

PART II

ITEM 5. MARKET FOR REGISTRANT’S COMMON EQUITY, RELATED STOCKHOLDER MATTERS AND ISSUER REPURCHASES OF EQUITY SECURITIES

Market for Registrant’s Common Equity

Our common stock has traded on the Nasdaq National Market since our initial public offering on October 5, 2000 under the symbol “KOSN.” Prior to such time, there was no public market for our common stock. The following table shows the high and low sales prices per share of our common stock as reported on the Nasdaq National Market for the periods indicated:

As of February 28, 2005, there were approximately 100 record holders of our common stock. We have never declared or paid dividends on our capital stock. We currently expect to retain future earnings, if any, for use in the operation and expansion of our business and do not anticipate paying any cash dividends in the foreseeable future.

Use of Proceeds From Sales of Registered Securities

Our initial public offering of common stock was effected through a Registration Statement on Form S-1 (File No. 333-33372) that was declared effective by the SEC on October 4, 2000. All 5,750,000 shares of common stock offered, inclusive of 750,000 shares of common stock subject to the underwriters’ over-allotment option, were sold at a price per share of $14.00 for gross proceeds of approximately $80.5 million.

We paid a total of approximately $5.6 million in underwriting discounts and commissions, and other costs and expenses, other than underwriting discounts and commissions, were approximately $1.5 million. After deducting the underwriting discounts and commissions and the offering costs and expenses, we received net proceeds from our initial public offering of common stock of approximately $73.4 million.

As of December 31, 2004, all net offering proceeds have been applied with approximately $8.2 million used to purchase property and equipment and approximately $65.2 million used for general corporate purposes.

35

Table of Contents

KOSAN BIOSCIENCES

ITEM 6. SELECTED FINANCIAL DATA

The statement of operations data for each of the years ended December 31, 2004, 2003 and 2002, and the balance sheet data as of December 31, 2004 and 2003, have been derived from our audited financial statements included elsewhere in this Annual Report on Form 10-K that have been audited by Ernst & Young LLP, independent registered public accounting firm. We have derived the statement of operations data for the years ended December 31, 2001 and 2000, and the balance sheet data as of December 31, 2002, 2001 and 2000 from our audited financial statements that are not included in this Annual Report on Form 10-K. Our historical results are not necessarily indicative of results to be expected for any future period. The data presented below have been derived from financial statements that have been prepared in accordance with accounting principles generally accepted in the United States and should be read with our financial statements, including the notes, and with “Item 7. Management’s Discussion and Analysis of Financial Condition and Results of Operations” included elsewhere in this Annual Report on Form 10-K.

36

Table of Contents

Recent Accounting Pronouncements

In December 2004, the FASB issued SFAS No. 123 (revised 2004), “Share-Based Payment” (“SFAS 123R”). SFAS 123R requires that the compensation cost relating to share-based payment transactions be recognized in financial statements. The cost will be measured based on the fair value of the instruments issued. SFAS 123R covers a wide range of share-based compensation arrangements including stock options and employee stock purchase plans. SFAS 123R replaces SFAS 123 and supersedes APB Opinion No. 25 (“APB 25”). As originally issued in 1995, SFAS 123 established as preferable the fair-value-based method of accounting for share-based payment transactions with employees. However, that Statement permitted entities the option of continuing to apply the guidance in APB 25, as long as the footnotes to financial statements disclosed what net income would have been had the preferable fair-value-based method been used. We plan to adopt SFAS 123R in our fiscal quarter ending September 30, 2005. We expect that the adoption of SFAS 123R will have a significant impact on our financial statements in that fiscal quarter and future periods but we cannot reasonably estimate the impact of adoption because we expect certain assumptions that can significantly affect the calculation of the value share-based payments to employees to change in 2005.

In March 2004, the FASB issued Emerging Issues Task Force No 03-1, “The Meaning of Other-Than-Temporary Impairment and Its Application to Certain Investments”, or EITF 03-1, which determines the meaning of other-than-temporary impairment and its application to investments classified as either available-for-sale or held-to-maturity under Statement 115 (including individual securities and investments in mutual funds), and investments accounted for under the cost method or the equity method. We adopted EITF 03-1 in our fiscal quarter ended March 31, 2004.

ITEM 7. MANAGEMENT’S DISCUSSION AND ANALYSIS OF FINANCIAL CONDITION AND RESULTS OF OPERATIONS

The following discussion and analysis should be read with “Selected Financial Data” and our financial statements and notes included elsewhere in this Annual Report on Form 10-K. The discussion in this Annual Report on Form 10-K contains forward-looking statements that involve risks and uncertainties, such as statements of our plans, objectives, expectations and intentions. The cautionary statements made in this Annual Report on Form 10-K should be read as applying to all related forward-looking statements wherever they appear in this Annual Report on Form 10-K. Our actual results could differ materially from those discussed here. Factors that could cause or contribute to these differences include those discussed in “Item 1. Business—Risk Factors That May Affect Results of Operations and Financial Condition” and elsewhere in this Annual Report on Form 10-K.

We are a biotechnology company developing drug candidates from an important class of natural compounds known as polyketides. We have proprietary technologies for the manipulation of polyketides, a rich source of pharmaceutical

37

Table of Contents

KOSAN BIOSCIENCES

products. We use our platform technologies to develop product candidates that target large pharmaceutical markets. We have two first-in-class anticancer agents in Phase II and Phase Ib clinical trials, both of which have follow-on compounds in Phase I clinical trials. Following is the status of our product candidates.

We also have additional product candidates in the areas of gastrointestinal motility and infectious disease that are undergoing preclinical evaluation and several early stage product and research technology development programs largely based on polyketides.

We have incurred significant losses since our inception. As of December 31, 2004, our accumulated deficit was $101.2 million. We expect to incur additional operating losses over the next several years as we continue to develop our technologies and fund internal product research and development.

Critical Accounting Policies and Management Estimates

Our discussion and analysis of our financial condition and results of operations are based upon our financial statements, which have been prepared in accordance with U.S. generally accepted accounting principles. The preparation of these financial statements requires us to make estimates and judgments that affect the reported amounts of assets, liabilities, revenues and expenses, and related disclosure of contingent assets and liabilities. On an on-going basis, we evaluate our estimates, including those related to our research and development collaborations,

38

Table of Contents

investments, deferred tax assets and their related valuation allowance, financing operations and contingencies. We base our estimates on historical experience and on various other assumptions that we believe to be reasonable under the circumstances, the results of which form the basis for making judgments about the carrying values of assets and liabilities that are not readily apparent from other sources. Actual results may differ from these estimates.

We believe the following critical accounting policies, which have been reviewed by our Audit Committee, affect our more significant judgments and estimates used in the preparation of our financial statements.

Revenue Recognition

We recognize license and other up-front fees over the estimated research and development term of the agreement. If the agreement does not have a specified research and development term, we must apply judgment in determining the appropriate level of recognition. As of December 31, 2004, we had $15.4 million in deferred revenue, representing the unamortized balance of the $25.0 million initial fee received in connection with our collaboration with Roche. This initial fee is being amortized through the second half of 2009, the current estimated clinical development period. In December 2004, upon the commencement of a Phase I clinical trial of KOS-1584, our second-generation Epothilone D compound, we determined that the estimated clinical development period extended from the second half of 2007, which was initially estimated when we entered into the collaboration with Roche in 2002, to the second half of 2009. Our initial estimate of the clinical development period contemplated development of one product candidate, KOS-862. The change in our estimate has resulted in a further deferral of the unrecognized portion of the initial fee. Any changes in our estimate will result in either an acceleration or further deferral of the related revenue recognition.

Clinical Trial Accruals

Over the next year, the clinical trials of our epothilone and geldanamycin analog programs will significantly increase our development expenditures. Research and development expenditures are expensed as incurred. Our expenses related to clinical trials are based on estimates of the services received and efforts expended pursuant to contracts with multiple research institutions and clinical research organizations that conduct and manage clinical trials on our behalf. The financial terms of these agreements are established prior to the initiation of the related clinical trial, thus establishing the basis of our estimates. However, these terms may be subject to amendment due to changes in the scope and length of the related clinical trial. Expenses related to clinical trials generally are accrued based on the level of patient enrollment and activity according to the protocol. We monitor patient enrollment levels and related activity and adjust our estimates accordingly. The process by which we currently estimate our clinical trial accruals is consistent with prior periods and, in the past, there have not been significant adjustments of the actual results to our estimates.

Stock-Based Compensation

Stock-based compensation expense for options granted to non-employees has been determined as the fair value of the consideration received or the fair value of the equity instruments issued, whichever is more reliably measured. We recognized other stock-based compensation for non-employees of approximately $605,000 in 2004, $619,000 in 2003, and $1.4 million in 2002. In addition, assuming no changes, we expect to recognize other stock-based compensation in connection with stock options granted to non-employees of approximately $512,000 in 2005, $153,000 in 2006, $119,000 in 2007 and $22,000 in 2008. The measurement of stock-based compensation to non-employees is subject to periodic adjustment as the underlying securities vest. As such, changes to these future period measurements could be substantial should we experience significant changes in our stock price. For example, a 50% increase in the price of our stock, from that of December 31, 2004, would translate into an approximately 66% increase in the related estimated expenses above. See Notes 1 and 9 of our financial statements.

39

Table of Contents

KOSAN BIOSCIENCES

In December 2004, the Financial Accounting Standards Board, or FASB, issued Statement No. 123 (revised 2004), “Share-Based Payment,” or SFAS 123R, which will require us to record a charge to earnings for the compensation cost relating to share-based payment transactions, including stock options and employee stock purchase plans, effective July 1, 2005. We are currently evaluating option valuation methodologies and assumptions permitted by the FASB for purposes of implementing the change in accounting treatment. This change will have a negative impact on our financial results. For an illustration of the effect of such a change in our recent financial results, see Note 1 to the Notes to Financial Statements—Summary of Significant Accounting Policies: Stock Options Plans. The adoption of SFAS 123R’s fair value method will have a significant impact on our results of operations, although it will have no impact on our overall financial position. The amounts related to SFAS 123R cannot be predicted at this time because it will depend will vary due to certain factors, including but not limited to, the number of new option grants and purchases made under the employee stock purchase plan, employee terminations, and changes in our stock price, and related assumptions. We plan to adopt SFAS 123R on July 1, 2005.

Results of Operations

Years Ended December 31, 2004, 2003 and 2002

Revenue

Revenues for the years ended December 31, 2004, 2003 and 2002 were approximately $22.9 million, $31.4 million and $9.6 million, respectively. Revenues in 2004 consisted primarily of contract revenue recognized under our development and commercialization agreement with Roche and funded research related to government grant awards. Revenues in 2003 and 2002 consisted primarily of contract revenue recognized under our development and commercialization agreement with Roche, our former collaboration agreement with Johnson & Johnson Pharmaceutical Research and Development LLC, or J&JPRD, and funded research related to government grant awards.

The decrease in revenues of approximately 27%, or $8.5 million, in 2004 compared to 2003 was the result of the following:

40

Table of Contents

The increase in revenues of approximately 227%, or $21.8 million, in 2003 compared to 2002 was the result of the following:

If we do not maintain or extend our agreement with our Roche, our revenues will significantly decrease unless we enter into additional collaborations that provide substantial revenues.

Research and Development Expenses

For the years ended December 31, 2004, 2003 and 2002, our research and development expenses were approximately $40.2 million, $36.8 million and $28.4 million, respectively. Our research and development activities consist primarily of salaries and other personnel-related expenses, clinical trial-related services performed by clinical research organizations and research institutions and other outside service providers, licensing-related expenses, lab consumables and facility-related expenses. We group these activities into two major categories: “research and preclinical” and “clinical development.”

The costs associated with research and preclinical and clinical development activities approximated the following:

The increase of 9%, or approximately $3.4 million, in research and development expenses for the year ended December 31, 2004 compared to the same period in 2003 was primarily due to the following:

41

Table of Contents

KOSAN BIOSCIENCES

The increase of 30%, or approximately $8.4 million, in research and development expenses for the year ended December 31, 2003 compared to the same period in 2002 was primarily due to the following:

We allocate salary-driven and space-use-driven overhead expenses to research and development and to general and administrative expenses based on salaries and space-utilization by each respective area. Our research and development employees decreased to 109 in 2004 from 110 in 2003. In 2004, we experienced increased salary and personnel-related expenses resulting from a controlled shift from research-skilled employees to more development-skilled employees as we continue to advance our product candidates into clinical development. We expect our research and development expenses will increase substantially as KOS-862, KOS-1584, KOS-953 and KOS-1022 advance further into the clinic, and as we advance our research programs into later stages of development.

The table below summarizes the current development status of our most advanced product candidates:

42

Table of Contents

Phase I clinical trials generally are expected to last between 12 and 18 months, Phase II clinical trials are expected to last between 18 and 24 months and Phase III clinical trials are expected to last between 24 and 42 months. The length of clinical development depends on the specific disease and patient population. For drug candidates that are in preclinical development, the timing of an IND filing varies significantly and is difficult to predict and therefore not reflected in the table above.

Roche is responsible for the majority of the costs of the epothilone clinical trials conducted under our collaboration and is funding all of the KOS-862 and KOS-1584 clinical trials other than the KOS-862 prostate cancer trial, which we are currently funding. We are responsible at our cost to supply 17-AAG for clinical trials sponsored by the NCI under the CRADAs, and the NCI is responsible for the remainder of the costs of these trials through Phase II. In addition, we are sponsoring other clinical trials of KOS-953 and plan to sponsor KOS-1022 clinical trials at our sole expense. Our research and development expenses do not reflect the costs incurred by our partners, Roche or the NCI, associated with the clinical trials they are conducting in connection with our epothilone and Hsp90 inhibitor programs, respectively.

General and Administrative Expenses

For the years ended December 31, 2004, 2003 and 2002, general and administrative expenses were approximately $5.9 million, $5.1 million and $4.9 million, respectively.

The increase of 16%, or approximately $797,000, in general and administrative expenses in 2004 compared to 2003 was primarily due to the following:

The increase of 4%, or approximately $205,000, in general and administrative expenses in 2003 compared to 2002 was primarily due to the following:

43

Table of Contents

KOSAN BIOSCIENCES

The number of our general and administrative employees increased to 25 at December 31, 2004 as compared to 22 at December 31, 2003. We expect our general and administrative expenses will increase in the future to support the continued growth of our research and development efforts.

Interest Income, Interest Expense and Gain on the Recovery of Investment

Interest Income. Interest income increased to approximately $1.4 million in 2004 from approximately $1.2 million in 2003. This increase was primarily due to higher average investment balances that resulted from our registered direct offering in December 2003. Interest income decreased to approximately $1.2 million in 2003 from approximately $2.2 million in 2002. This decrease was due to lower average investment balances and lower investment yields associated with the declining interest rate environment in 2003.

Interest Expense. Interest expense increased to approximately $343,000 in 2004 from approximately $337,000 in 2003. This increase was due to additional equipment debt financing and an increasing rate environment during 2004, which resulted in a higher average debt balance at higher average interest rates. Interest expense decreased to $337,000 in 2003 from $356,000 in 2002. This decrease resulted from the scheduled repayment of existing debt at higher average interest rates, partially offset by additional equipment debt financing at lower average interest rates in 2003. We expect our interest expense will increase in the future resulting from additional property and equipment-related debt financings in the increasing interest rate environment.

Gain on the Recovery of Investment. At June 30, 2001, we determined that an other-than-temporary decline in the fair value of our investment in short-term commercial paper of Southern California Edison, a utility company, had occurred. Accordingly, we recorded approximately $990,000 loss on the related write-down in the carrying value of the investment. In March 2002, the security was fully repaid. Thus, we recorded a realized gain on the investment of approximately $990,000 for the year ended December 31, 2002.

Provision for Income Taxes

We incurred net operating losses in the years ended December 31, 2004, 2003 and 2002 and consequently did not pay federal or state income taxes. As of December 31, 2004, we had federal net operating loss carryforwards of approximately $65.3 million. We also had federal research and development tax credit carryforwards of approximately $2.6 million. If not utilized, the net operating losses and credit carryforwards will expire at various dates beginning in 2010 through 2024. As of December 31, 2004, we had state net operating loss carryforwards and state research and development tax credit carryforwards of approximately $36.8 million and $2.7 million, resprectively. The state net operating loss carryforwards will expire at various dates beginning in the year 2005 through 2014, if not utilized. The state research and development carryforwards do not expire. Use of the net operating losses and credits may be subject to a substantial annual limitation due to the change in ownership provisions of the Internal Revenue Code of 1986, as amended, and similar state provisions. The annual limitation may result in the expiration of net operating losses and credits before utilization. See Note 11 of our financial statements.

44

Table of Contents

Liquidity and Capital Resources

Since inception we have financed our operations primarily through sales of our convertible preferred stock and common stock, contract payments received under our corporate collaboration agreements and government grant awards, interest income and equipment financing arrangements. As of December 31, 2004, we had received approximately $149.7 million from the sales of convertible preferred and common stock, approximately $91.8 million from contract payments received under our corporate collaboration agreements and government grant awards, approximately $14.1 million from interest income and approximately $11.7 million from equipment financing arrangements. As of December 31, 2004, we had approximately $83.4 million in cash and investments, compared to approximately $105.3 million as of December 31, 2003. Our funds are currently invested in U.S. Treasury and government agency obligations and corporate obligations.

Cash used in operating activities was approximately $18.2 million in 2004, compared to cash provided by operations of approximately $1.9 million in 2003. Our net loss of approximately $22.1 million in 2004 was partially offset by non-cash expenses of approximately $5.0 million related to stock-based compensation, depreciation and amortization of property and equipment and investment premiums and discounts, partially offset by an approximately $1.0 million decrease in assets and liabilities primarily due to recognition of deferred revenue related to the upfront payment from Roche, partially offset by an increase in receivable collections from Roche, net of payments related to our settlement agreement with Sloan-Kettering. We do not anticipate generating cash from operating activities for the next several years. Our net loss of approximately $9.7 million in 2003 was offset by the receipt of the second of two $12.5 million installments of an initial fee from Roche and non-cash expenses of approximately $4.6 million related to stock-based compensation, depreciation and amortization of property and equipment investment premiums and discounts, partially offset by an approximately $5.1 million increase in accounts receivable primarily due to contract revenue and a milestone related to our Roche collaboration.

Our investing activities, excluding changes in our investments, for the year ended December 31, 2004, used cash of approximately $2.3 million, compared to approximately $4.3 million in 2003, reflecting approximately $1.9 million related to the purchase of additional laboratory and office equipment and approximately $354,000 related to the renovation of our facilities. Investing activities in 2003, excluding changes in our investments, reflected approximately $1.9 million related to the renovation and expansion of our facilities and approximately $2.4 million related to the purchase of additional laboratory and office equipment. We expect to continue our purchases of equipment and leasehold improvements in 2005 related to the enhancement and expansion of our leased facilities.

Cash provided by financing activities was approximately $729,000 for the year ended December 31, 2004, compared to approximately $28.6 million in 2003. Financing activities in 2004 included approximately $1.3 million of equipment debt financing and $1.5 million in proceeds from the sale of our common stock through stock option exercises and stock purchases made under our 2000 Employee Stock Purchase Plan, offset by $2.1 million of scheduled payments on new and existing debt. Financing activities in 2003 included approximately $26.0 million net proceeds from the sale of our common stock in a registered direct offering in December 2003, $3.1 million of equipment financing, net of scheduled payments on new and existing debt, and $776,000 related to proceeds from the repayment of notes receivable from stockholders.

In April 2004, we entered into a $3.5 million equipment line of credit agreement for facility improvements and capital purchases, which expires in April 2005. As of December 31, 2004, we had utilized approximately $800,000 of the line of credit leaving approximately $2.7 million available for future draws.

We believe that our existing cash and investment securities and anticipated cash flow from our existing collaboration with Roche will be sufficient to support our current operating plan into the second half of 2006. We may, however,

45

Table of Contents

KOSAN BIOSCIENCES

choose to obtain additional financing from time to time, or focus our spending on priority programs to provide for the proper alignment of resources and maintenance of sufficient levels of financial reserves. We have based this estimate on assumptions that may prove to be wrong. Our future capital uses and requirements depend on numerous forward-looking factors, including the following: