Item 1. Business

Overview of Array's Business

        We are a drug discovery company creating new small molecule drugs through the integration of chemistry, biology and informatics. Our experienced scientists use the Array Discovery Platform, our integrated set of drug discovery technologies, to invent novel small molecule drugs in collaboration with leading pharmaceutical and biotechnology companies and to build our own pipeline of proprietary drug candidates.

        The drug industry is experiencing revolutionary change fueled by genomics and by the tremendous progress in the biological understanding of disease. Historically, a key bottleneck in the development of new drugs has been the identification of targets, which are proteins that may play a role in disease. However, recent advances in genomics and biology have resulted in the identification of thousands of new potential targets. As a result of the proliferation of new targets, we believe the drug research and development bottleneck is shifting from the identification of new targets to the creation of safe and effective new small molecule or protein-based therapeutics.

        Small molecule drugs are invented by chemists and are generally taken as a pill, as opposed to protein-based therapeutics which are generally given by injection. We believe small molecule drugs have inherent advantages over protein-based therapeutics, including a greater universe of treatable diseases, lower cost with greater ease of manufacturing, and patient preference for a pill over an injection. Although a high proportion of biotechnology research has historically been devoted to protein-based therapeutics, approximately 88% of the top 200 prescription drugs, based on worldwide sales in 2001, are small molecule drugs. Accordingly, we believe that there will be increased emphasis on small molecule drug discovery in the biotechnology industry.

        Our aim is to be the industry leader in small molecule drug discovery by utilizing the Array Discovery Platform to efficiently create high-quality drug candidates. Early in the drug discovery process, our scientists use the Array Discovery Platform to engineer into a drug candidate desirable drug characteristics, such as improved potency, specificity and dosing regimen and reduced side effect profile. We believe that the early optimization of superior drug characteristics will reduce the failure rate of drug candidates, thus increasing research productivity.

        To capitalize upon opportunities in drug discovery, we believe that an experienced scientific team with a track record of success in drug discovery is crucial. Accordingly, we have grown our staff to 245 full-time employees as of June 30, 2002, including 179 scientists, of whom 110 have Ph.D.s and 85 have experience at large pharmaceutical or biotechnology companies. Members of our scientific staff have contributed during their careers to more than 20 Investigational New Drug applications, or INDs, and are inventors on over 200 drug-related patents and patent applications and are authors on over 1,100 scientific publications. Additionally, we have increased the laboratory space necessary for our continued growth by securing long-term leases, which will provide a total of 219,000 square feet over the next two years to accommodate up to 350 scientists.

         Achievements. Our fiscal 2002 achievements include:

•

Increasing revenues by 106% to $35.1 million compared to $17.0 million in fiscal year 2001;



•

Achieving positive EBITDA of $997,000 for fiscal year 2002, compared to a negative EBITDA of $4.7 million in fiscal year 2001, realizing a $5.7 million improvement. EBITDA is calculated as net loss as reported, minus net interest income, plus depreciation, plus compensation related to stock option grants;

11

•

In collaboration with ICOS Corporation, our first drug discovery agreement resulted in a clinical candidate. In November 2001, ICOS initiated a clinical trial with IC485 and subsequently made a milestone payment to us;



•

Initiating success-based collaboration agreements, which include research funding and potential milestone and/or royalty payments, with the following companies: Amgen Inc.; Aptus Pharmaceuticals; ICOS; Japan Tobacco Inc.; Takeda Chemical Industries, Inc.; Syrrx, Inc.; Trimeris, Inc.; and Vertex Pharmaceuticals Incorporated;



•

Tripling the investment in Array's proprietary research to $5.5 million, compared to $1.6 million in the prior year. We advanced three internal drug discovery programs into lead optimization. These programs are focused on the kinase family of protein targets for inflammation and oncology indications;



•

Expanding our access to novel targets for Array's proprietary research through a joint development program with Callisto Pharmaceuticals Inc. The program brings together Callisto's expertise in identifying important targets and the Array Discovery Platform. The parties will jointly own and fund the research and development of products that result from this agreement; and



•

Completing construction of our current Good Manufacturing Practices (cGMP) manufacturing facility, which will allow us to produce cGMP compliant compounds for Phase I clinical testing. We are validating this capability for compliance with FDA regulations and anticipate being able to initiate our first cGMP manufacturing campaign in the second half of calendar 2002.

         Strategy. Our objective is to build the industry's premier drug discovery company by:

•

Continuing to enhance the Array Discovery Platform by developing novel tools and implementing new technologies to be the industry's most efficient inventor of small molecule drug candidates;



•

Identifying drug candidates in collaboration with leading pharmaceutical and biotechnology companies, where we receive research funding and share in the success we create through potential milestone and royalty payments; and



•

Using the Array Discovery Platform to build our own pipeline of proprietary drug candidates, which we intend to continue to license for co-development and commercialization with partners.

Drug Discovery and Development

        Drug discovery and development is the process of creating and evaluating drugs for the safe and effective treatment of human disease. Today, this process requires biological, chemical and informatics expertise. The role of biology in drug discovery includes understanding the mechanism of diseases, identifying potential targets for therapeutic intervention and evaluating potential drug candidates. The role of chemistry in drug discovery is the invention of safe and effective new chemical entities, or drug candidates, to address these targets. The role of informatics in drug discovery is to improve decision-making by identifying and replicating the characteristics of successful drugs, efficiently sharing current knowledge and creating databases to predict future clinical success. Drug discovery and development comprises a number of interrelated disciplines.

        Target identification.     Targets are proteins that may play a fundamental role in the onset or progression of a particular disease. Biologists identify targets against which chemists create drug candidates. Until recently, pharmaceutical researchers were limited to studying approximately 500 biological targets. The number of available biological targets is being vastly expanded through genomics. Pharmaceutical and biotechnology companies are advancing many of these newly identified

12

potential targets into drug discovery. Many other potential targets have yet to be validated, meaning that their roles in causing disease are not completely understood.

        Drug discovery.     Drug discovery includes structural biology, lead generation, lead optimization and process research and development.

•

Structural biology. Structural biology is the process of cloning, expressing and purifying protein targets to create information about their functions and how they interact with drug candidates.



•

Lead generation. Lead generation is the process of identifying lead molecules that interact with a target with sufficient potency and selectivity to warrant further testing and refinement as possible drug candidates. During lead generation, researchers develop tests, called assays, to screen libraries, or large collections of potential lead compounds, against targets to evaluate their therapeutic value.



•

Lead optimization. Lead optimization is the complex, iterative process of refining the chemical structure of a compound to improve its drug characteristics with the goal of producing a preclinical drug candidate. Researchers focus on a number of considerations in optimizing a drug candidate, including the following drug characteristics:

•

Process research and development. In the earlier stages of drug discovery, small, typically milligram quantities of the compound are used. Before a drug candidate can be taken into clinical trials, larger, typically kilogram quantities often must be synthesized. The goal of process research is to streamline the synthesis of larger quantities of the compound by minimizing the number of synthetic steps chemists must perform and reducing the time and cost of production. Once a successful process is developed, batches of compounds produced in accordance with cGMP processes are synthesized for animal and human testing in the preclinical and clinical development phases.

        Preclinical development.     Prior to human clinical testing, a potential drug candidate must undergo extensive in vitro, or laboratory, and in vivo, or animal model, studies to predict human drug safety. These studies investigate toxicity over a wide range of doses and the mechanism by which the drug is metabolized. The objective of preclinical testing is to obtain results that will allow a drug candidate to enter human clinical trials through approval of an IND application by the Food and Drug Administration.

        Clinical development.     Clinical trials, or human tests to determine the safety and efficacy of potential drug candidates, are typically conducted in three sequential phases, although the phases may overlap. Successful clinical trials will result in the filing of a New Drug Application, or NDA, with the FDA in order to obtain approval to market the drug in the United States. Similarly, clinical trials must be conducted and regulatory approvals secured before a drug can be marketed in other countries.

13

Issues and Opportunities in Drug Discovery

        Inefficiencies in drug discovery and development.     Despite all of the recent technological advances and investment in genomics, biology, chemistry and informatics, drug research and development remains slow, expensive and risky. It is estimated that from 1991 to 2001 annual research and development spending more than tripled, while the number of new drugs approved by the FDA during the ten-year period ended in 2001 only increased by 35% over the ten-year period ended in 1991. We believe this disparity between the increase in research and development spending and the significantly lower increase in approved new drugs indicates the need for improved productivity in the drug discovery process.

        The challenge of turning genomics information from targets into drugs.     The research and development process is experiencing a fundamental change fueled by the revolution in genomics, which has resulted in the identification of many new targets. We believe the resulting proliferation of targets from 500 to between 3,000 and 10,000 has shifted the bottleneck in drug research and development from the identification of new targets to the creation of safe and effective drugs for these targets. We believe the knowledge gained from genomics and biology and the resulting drug discovery bottleneck, coupled with the advantages of small molecule drugs, will lead to a dramatic increase in the investment in small molecule drug discovery.

        Importance of chemistry in the drug discovery process.     The chemical make-up or structure of a drug is the key determinant of its potency, specificity, dosing regimen and side effect profile. Minor modifications in chemical structure can differentiate drugs and determine their success or failure in the marketplace. While targets are used for evaluating the drug characteristics of chemical compounds, chemistry is necessary to invent the composition of the actual drug. Therefore, while the ultimate value of intellectual property associated with newly identified targets is currently unknown, the value of intellectual property associated with drugs invented by chemists is known to be significant.

         Need for additional capacity and improved research productivity within the pharmaceutical industry. The demands for new and improved drugs, coupled with the emerging potential of new targets, have increased competition for qualified chemists. A number of pharmaceutical companies have revealed plans to increase their discovery chemistry capacity over the next five years. Consequently, we believe competition for qualified chemists to fill these positions will continue.

        Need for drug discovery capabilities within the biotechnology industry.     Many biotechnology companies are increasing their focus on creating drugs against their proprietary targets. Historically, they have partnered with pharmaceutical companies to create small molecule drugs. These arrangements have often resulted in biotechnology companies relinquishing much of the potential economic value resulting from their discoveries. Accordingly, several biotechnology companies have announced their intention to build small molecule drug discovery capabilities internally or through acquisitions. We believe they face significant barriers in creating a competitive drug discovery platform, which include: the difficulty of hiring multidisciplinary teams of scientists with drug discovery experience; the significant investment necessary to build and equip specialized laboratories; the difficulty in identifying and integrating acquisition opportunities; and, most importantly, the opportunity cost in time required to build an effective drug discovery capability.

The Array Solution

        Our solution to address the issues and opportunities in drug discovery is the Array Discovery Platform, our integrated suite of drug discovery technologies, which is designed to increase research productivity through creating higher quality drug candidates and lowering clinical attrition rates. This platform includes the pragmatic integration of appropriate drug discovery technologies, enabling research tools and knowledge management through an electronic notebook and predictive

14

computational modeling. We utilize predictive computational modeling to improve experimental design and predict favorable drug characteristics. Our experienced scientists focus on data quality, not quantity, creating superior drug candidates through improved decision-making. We believe we have implemented a unique solution to bridge the gap between target discovery and clinical development and to address the issues and opportunities in drug discovery by:

•

Hiring experienced scientists and organizing them into multidisciplinary teams;



•

Enhancing the Array Discovery Platform by continuing to develop, acquire and integrate appropriate new technologies that emphasize high quality data generation;



•

Creating and utilizing enabling research tools to accelerate the execution of experiments;



•

Building knowledge management tools to improve the design of experiments and predict favorable drug characteristics; and



•

Identifying lead compounds and designing drug candidates against multiple targets within protein families in a parallel fashion.

        We have assembled a scientific team with experience in both the pharmaceutical and biotechnology industries to implement our solutions. During their careers, these scientists have contributed to more than 20 INDs, and have been inventors on over 200 drug-related patents and patent applications and authors on over 1,100 scientific publications.

Strategy

        Our objective is to become the leading inventor of high-quality drug candidates by building the Array Discovery Platform into the industry's premier drug discovery capability. Our strategies to achieve this objective are as follows:

        Accelerate the creation of high-quality drug candidates.     Our integrated drug discovery approach simultaneously leverages multiple technologies within the Array Discovery Platform to enable our scientists to share knowledge and is designed to improve decision-making across the organization. Our experienced scientists utilize the Array Discovery Platform to invent drug candidates by understanding the complex relationships between chemical structure and desirable drug characteristics. We believe this approach speeds the creation of high-quality drug candidates.

        Become the drug discovery partner of choice.     We provide collaborators with a fully integrated drug discovery capability to create drug candidates against their targets. While collaborators can access individual components of this capability, we intend to continue expanding collaborations across the entire Array Discovery Platform to become the drug discovery partner of choice.

        Invent our own drug candidates.     Our investment in our own proprietary research is intended to enhance the long-term value of our company. We use the Array Discovery Platform to invent our own proprietary drug candidates against known, validated targets and expect to increase our investment in our proprietary research efforts. Our focus is on improving current drug therapies for which we have identified deficiencies. We intend to commercialize these drug candidates by entering into collaborations to co-develop and commercialize these drug candidates with pharmaceutical and biotechnology partners.

        Create a world-class scientific research environment.     We expect to grow our business by continuing to aggressively recruit experienced scientists. Our continued success in recruiting and retaining these scientists depends upon the maintenance of our culture, which emphasizes quality science, innovation and empowerment of our scientists, and our ability to provide industry competitive salaries and equity participation in our company. We are committed to continuous process improvement, implementation of new technologies, shared learning among our scientists and innovative organizational design.

15

        Expand our capabilities through internal development and acquisitions.     We intend to acquire or develop new technologies and capabilities to expand the Array Discovery Platform. We have completed construction of a cGMP manufacturing facility, which will allow us to produce chemical compounds that meet cGMP requirements in quantities for Phase I clinical trials. We are validating this capability for compliance with FDA regulations and anticipate being able to initiate our first cGMP manufacturing campaign in the second half of calendar 2002. In the future, we may acquire additional laboratory sites to meet future needs and better attract scientific talent.

The Array Discovery Platform

        The Array Discovery Platform includes the following capabilities:

        Structural biology.     Our experienced biology teams are creating a better understanding of how small molecule drugs interact with targets. These teams clone, express and purify related families of protein targets across multiple therapeutic areas to gain insights into their function. X-ray crystallography and computational modeling are used to define the three-dimensional structures of these proteins. This process provides valuable information about the interactions between leads and targets.

        High throughput screening.     We develop our own assays or format assays supplied by a collaborator for high throughput screens and can screen up to 100,000 compounds per week. These assays are then used to screen against tens of thousands of small molecule compounds to obtain quantitative measures of drug quality. We also screen selected compounds against metabolism and toxicology assays both to establish quality and to populate our predictive databases. Our computational and medicinal chemists then mine this information to design focused libraries of small molecule drug candidates. This approach is designed to accelerate the discovery process by taking advantage of the similarities within target families.

        Predictive informatics.     Predicting drug characteristics, such as potency, dosing frequency and potential side effects, requires powerful data mining and management tools. Our informatics teams comprise computational chemistry, scientific computing and medicinal chemistry experts who work together to increase the probability of creating a successful drug. Our proprietary software enables our scientists to search databases of existing drugs, to generate novel predictive databases and to create modeling programs designed to better forecast drug characteristics. In addition, we use an electronic notebook to allow our scientists to collect and access information directly in the laboratory and throughout the organization. We believe the integration of these technologies improves scientific decision-making, resulting in higher quality drug candidates.

        Lead generation.     Our lead generation teams create and identify chemical compounds that demonstrate desirable drug characteristics when screened against a target. Compounds that warrant further testing and refinement as potential drug candidates are called leads.

    Optimer building blocks

        We believe that chemists can create high value compounds more rapidly by using quality building blocks and automated chemical synthesis techniques. We recognize that a constraint in drug discovery is the availability of these high quality building blocks. Our chemists have used our proprietary software, RADICAL, and their experience in assessing drug-relevant chemical structures to design a series of building blocks with desirable drug-relevant properties. These building blocks are added to a core chemical structure, or a scaffold, during compound synthesis and are an important component of our overall drug discovery strategy. We produce primary building block sets for construction of our Lead Generation Libraries. We then use sets of complementary secondary building blocks for creating focused libraries to determine structure activity relationships, or SARs, in lead optimization programs.

16

    Lead Generation Libraries

        A critical rate-limiting step in the drug discovery process is the availability of high-quality compound libraries that have been designed for screening specifically against important target classes and for subsequent rapid lead optimization. We believe that the production of large compound libraries, by itself, has limited value for creating high-quality leads. Instead, we design our libraries so that any leads require less optimization. We believe this approach will result in clinical candidates with a greater likelihood of clinical success.

        We design our libraries according to the following criteria:

•

Biologically-relevant diversity. We have established specific computational parameters to define the diversity of our compound libraries. Our proprietary informatics tools categorize how changes in chemical structure correlate with the biological activity of known drugs and use this information to define our diversity parameters. Libraries can be constructed to optimize diversity and therefore maximize the information provided by each library compound.



•

Capture full patent potential. Our scientists maximize the number of distinct three-dimensional drug-like shapes, or pharmacophores, during library design. This approach is designed to optimize the number of discrete, patentable compound sets within a library, with the goal of identifying the key structural features of drug-target interactions.



•

Target-directed chemical scaffolds. Our chemists create scaffolds designed for disease-related families of targets. We believe this scaffold strategy allows us to increase the probability of finding a high-quality lead for a given target. We attach our novel building blocks to these scaffolds to create our library compounds.



•

Drug-relevant building blocks. We use drug-relevant building blocks to synthesize libraries. We design the library to identify the least complex structure that will interact with a target. Any lead generated from our Lead Generation Libraries can be readily optimized through the use of more complex building block sets. These focused libraries provide initial SAR around any lead.



•

Optimized chemical synthetic processes for high purity. We invest significant effort in the process design and synthesis of each library to ensure that the compounds generated are highly pure and can be readily optimized. The library undergoes analysis during each stage of its development to ensure the identity of each compound and maintain overall quality.

        We provide chemical compounds from our Lead Generation Libraries to our collaborators under a non-exclusive license for internal research. We also synthesize custom libraries, which we typically offer on an exclusive basis to individual collaborators, focused on specific target families or our collaborators' proprietary scaffolds. We retain all other rights to the compounds in our Lead Generation Libraries, including the right to use these compounds for our internal and collaborative programs, as well as the rights to the synthetic processes used to create these compounds. We create sub-libraries that interact with specific target families, including G-protein coupled receptors, nuclear receptors, enzymes and protein-protein interactions. The majority of all drugs on the market today are aimed at targets within these families.

        Analytical chemistry.     High-purity compounds are critical to the success of high throughput screening strategies in lead generation. In our experience, low-purity compounds result in a higher proportion of false leads, which waste discovery resources. Our analytical chemistry teams use automated instrumentation to evaluate the purity of chemical compounds, analyze the chemical processes used to synthesize these compounds and measure important drug properties. This capability allows for the high throughput analysis and purification of thousands of compounds per week.

        Lead optimization.     Leads that interact with targets may come from several sources, including our libraries, rational drug design, scientific literature and our collaborations or joint ventures. Regardless

17

of the source, we apply the same defined processes to optimize these leads to clinical drug candidates. We first utilize information regarding the three-dimensional structure of the target-lead interaction to design novel sets of compounds with the potential for better potency for synthesis. Next, we use our informatics capability to eliminate certain compounds that are predicted to have poor drug characteristics. We then synthesize, analyze and purify this refined set in a parallel format and screen these compounds in select assays to quantify drug characteristics. An iterative process of making small changes in chemical structure, evaluating the results and engineering improvements into the drug candidate is used to optimize its interaction with a target and refine its drug characteristics.

        Drug metabolism.     When optimizing desirable drug characteristics, it is often critical to determine how drugs are modified by the body at an early stage in the discovery process. We have established a series of assays to identify these metabolic changes. These assays include human liver enzyme assays, cellular assays and assays based upon fluids obtained from animals dosed with the compounds. We measure both the rate at which compounds are metabolized and how they are metabolized using mass spectrometry techniques. We also screen selected compounds in these assays to build drug metabolism databases to help predict clinical success of our future compounds.

        Process research and development.     Our process chemists improve complex synthetic procedures to allow for more efficient scale-up and production of drug candidates. We design proprietary processes to lower the cost and increase the rate at which drug candidates can be synthesized. We believe the experience of our process chemists in resolving complex synthetic problems allows us to rapidly develop new synthetic procedures and to accelerate the development of valuable drug candidates for human testing. Our goal is to apply these skills and this experience to create novel yet efficient processes to synthesize complex molecules.

•

Process design and scale-up. Once a potential drug candidate has been identified, it is critical to reach a rapid decision whether to advance the candidate into the clinic. In many cases, lack of an adequate quantity of a specific compound for preclinical testing delays that decision. Our efforts reduce the number of steps in complex medicinal chemistry processes and improve yields to allow for the rapid synthesis and scale-up of preclinical and clinical drug candidates.



•

Custom synthesis. Our chemists can undertake challenging syntheses on a custom basis to produce building blocks, complex intermediates and final products in either small scale or bulk quantities. We synthesize compounds both on a proprietary and non-proprietary basis. We have synthesized for a number of collaborators increasingly larger quantities of compounds, to meet their research needs. We intend to create proprietary processes that can be licensed to collaborators as they advance potential drug candidates into clinical trials. We have the capacity to produce lots of up to 10 kilograms.



•

cGMP manufacturing . Array's chemists are skilled in rapidly producing the first qualified Active Pharmaceutical Ingredient lot with the goal of ultimately reducing a collaborator's timeline for entering clinical trials. We have completed construction of a cGMP manufacturing facility, which will allow us to produce chemical compounds that meet cGMP requirements for Phase I clinical testing. We are validating this capability for compliance with FDA regulations and anticipate being able to initiate our first cGMP manufacturing campaign in the second half of calendar 2002. This capability features a facility with three cGMP kilo-labs, including 100-liter equipment, and will allow us to manufacture and qualify analytical reference standards and impurities.

        We believe the Array Discovery Platform, our integrated approach to drug discovery, enables both our collaborators and our internal discovery teams to create higher quality drugs more quickly and less expensively.

18

Proprietary Drug Discovery

        We use the Array Discovery Platform to invent our own proprietary drug candidates. We plan to co-develop and commercialize these drug candidates in partnership with pharmaceutical and biotechnology companies. Our pragmatic chemical-proteomics approach is designed to increase research productivity by taking advantage of the similarities in drug design strategies for related targets within a protein family. We identify and access disease-associated targets through publicly available human genomics information and scientific literature as well as through collaborations and joint ventures. These protein families will typically have similar three-dimensional structures and related biological function and/or enzymatic activity. More importantly, the experimental design expertise required to create drugs against a given target is captured and replicated against the entire target family. In parallel, we synthesize focused compound libraries, which are designed to interact with targets within a family. By screening these libraries against several targets within a family, we seek to generate multiple leads with desirable drug characteristics. Our scientists then optimize the drug characteristics of these leads to provide clinical development candidates.

        Our multidisciplinary project teams focus on biologic functions, or pathways, that have been validated as important to the treatment of human disease. We are currently working on a number of target families, including cytokine receptors, kinases, proteases and G-protein coupled receptors, which are important targets for the treatment of arthritis, diabetes, Alzheimer's disease and cancer. We have identified proprietary lead series against certain kinases that regulate Tumor Necrosis Factor, or TNF, biosynthesis useful for the treatment of rheumatoid arthritis. We have also identified proprietary lead series against certain kinases that regulate cellular proliferation for the treatment of cancer. In addition, we have proprietary X-ray crystallographic structural data involving a small molecule bound to the Interleukin 1 receptor. This information may facilitate the creation of new oral treatments for arthritis. In each case, drugs that affect these targets have shown efficacy in human clinical trials or animal models of human disease. However, they have significant deficiencies related to mode of delivery, dosing frequency or side effect profile that may limit their use.

        Once we have qualified a valuable lead through secondary screening, or have created a significant intellectual property position, we will seek to maximize its risk-adjusted value by advancing the research program independently as far as Phase I clinical testing or initiating collaborations earlier with partners for subsequent lead optimization, co-development and commercialization. Our research efforts on specific phosphatase target family members resulted in collaboration agreements with Amgen and Vertex providing an up-front fee, license fees, research funding and/or milestone payments and royalties. At the end of 2001, we regained the rights to the Amgen program while initiating a new research program with Amgen. We expect to enter into future collaborations that may provide similar fees and allow us to participate in the success of potential drug candidates through milestone and royalty payments. With other programs where value is especially high and research progress is rapid, we plan to move selected compounds into Phase I clinical development.

Commercialization

        We intend to maximize the value we capture by focusing our scientific resources on collaborations that use the full breadth of our capabilities and on our proprietary drug programs that enable us to participate in the success of the drug candidates that we create.

        Our intent is to increase revenue by continuing to expand our collaborations across the Array Discovery Platform. We enter into collaborations with pharmaceutical and biotechnology companies and receive fees for each scientist dedicated to these programs. In addition, in a number of our current collaboration agreements we are entitled to up-front fees, milestone payments upon achievement of certain drug discovery objectives and/or royalties based upon sales of products commercialized by our collaborators as a result of these agreements. We also sell or license research tools, including our

19

Optimer building blocks and our Lead Generation Libraries, on a non-exclusive basis to multiple collaborators, creating a recurring revenue stream.

        We create proprietary drug candidates with the intent of furthering their development and increasing their potential commercial value through collaborations with biotechnology or pharmaceutical partners. In the future, we may choose to advance certain drug candidates as far as early clinical development before entering into a collaboration agreement to maximize the value we retain. As we advance candidates, we will seek collaborations that provide us with an initial licensing fee for exclusive rights to our intellectual property, payments for continued research and down-stream payments that include milestone and/or royalty payments.

Our Collaborators

        A key element of our strategy is to increase the value we provide collaborators by expanding our relationships with them across complementary development efforts. Below, we describe our most significant collaborations:

        Amgen Inc.     In April 2000, Amgen began working with us by purchasing building blocks from our Optimer collection on a non-exclusive basis. In October 2000, we entered into a research and license agreement with Amgen. Under this agreement, we granted Amgen an exclusive license to one of our proprietary research programs and initiated joint research on potential drug candidates targeting PTP-1B, a target for diabetes. In November 2001, Amgen initiated a new drug discovery program with us, which replaced the PTP-1B program. We retained all rights to the existing PTP-1B program. Under the new program, Amgen will pay an up-front fee and fees based upon the number of our scientists working on the research phase of the agreement. We are also entitled to receive success payments based on the attainment of certain milestones.

        Eli Lilly and Company.     In March 2000, we entered into a research agreement with Eli Lilly to form a chemistry-based research collaboration. Under the terms of the agreement, up to 30 of our scientists will provide drug research in collaboration with Eli Lilly scientists on identified Eli Lilly drug discovery projects. We are compensated based upon an annual rate for each full-time equivalent employee working on an Eli Lilly project. Initially, this collaboration focused on certain aspects of our lead optimization chemistry. However, Eli Lilly has since expanded these joint efforts to other aspects of the Array Discovery Platform. Our agreement with Eli Lilly terminates in March 2005, but Eli Lilly may terminate the agreement at any time upon payment of an early termination fee.

        ICOS Corporation.     ICOS was our first drug discovery collaborator and has now taken advantage of the entire Array Discovery Platform. Our first agreement with ICOS, initiated in December 1998, addressed lead optimization of up to four ICOS targets. Under this agreement, our scientists, in collaboration with ICOS' scientists, developed clinical candidates from ICOS' preliminary leads. Based upon the success of this program, ICOS expanded our collaboration in the spring of 1999, by both initiating a second lead optimization program on a separate set of targets and subscribing to our Lead Generation Libraries. In less than one year, our initial collaboration led to the development of a clinical candidate, IC485, for a target called phosphodiesterase 4, or PDE4, for the treatment of inflammatory conditions. To speed the development of this clinical candidate, ICOS chose to access our chemistry process research to refine the production process to produce sufficient quantities for preclinical and early phase clinical testing. In November 2001, ICOS announced the initiation of a Phase I clinical trial for IC485, and we received a milestone payment for the achievement of this objective. We are entitled to additional milestone payments upon the achievement of specific clinical objectives.

        In July 2000, we consolidated and expanded our lead optimization agreements with ICOS into a drug discovery collaboration agreement for lead optimization on undisclosed targets. Under this

20

agreement, ICOS has the exclusive worldwide right to develop and market any products resulting from the collaboration. We are compensated based on an annual rate for each full-time equivalent employee working on an ICOS project and are entitled to milestone payments upon achievement of identified development and commercialization goals for products resulting from the collaboration. In March 2001, we expanded this lead optimization agreement and entered into a compound library agreement with ICOS. The agreement expires in July 2003, and may be terminated upon 90 days' notice by ICOS.

        In August 2001, we entered into an additional drug discovery collaboration agreement to discover and develop small molecule drugs directed at two specific targets containing the I-Domain allosteric site, or IDAS, structural motif. IDAS-targeted drugs regulate function of the target proteins through a novel allosteric mechanism. ICOS has identified key structural features of proteins containing the IDAS motif that will be exploited by our scientists to systematically produce drugs against targets of this class. Under the terms of this agreement, ICOS will provide us with research funding over two years. Our scientists and ICOS' scientists will collaborate in all aspects of lead generation and lead optimization. ICOS will be responsible for clinical development and commercialization. We are entitled to receive success payments upon reaching certain development milestones and royalties based upon sales of products resulting from this collaboration.

        Japan Tobacco Inc.     In March 2002, we entered into a drug discovery collaboration agreement with the pharmaceutical division of Japan Tobacco Inc. to create small molecule therapeutics against a proprietary Japan Tobacco target. Japan Tobacco will fund drug discovery research conducted by us based upon the number of Array scientists working on the research phase of the agreement. We will be entitled to receive milestone payments based on the selection and progress of clinical drug candidates, as well as royalties on net sales of products derived from the collaborative efforts.

        Merck and Co., Inc.     In May 1999, Merck purchased building blocks from our Optimer collection on a non-exclusive basis. This initial introduction led to an agreement for the exclusive development and supply of custom synthesized compounds for Merck. Building on this relationship, in September 2000, we announced an agreement with Merck for process research, synthesis and supply of custom libraries for Merck's drug discovery programs. We will develop processes for the synthesis of each library in collaboration with Merck scientists and utilize our proprietary high-speed synthesis and parallel purification platforms to create these high-quality libraries. Under the terms of the agreement, which expires in December of 2003, Merck will provide us with research funding as well as payment upon delivery of compounds.

        Pfizer Inc.     In October 2001, we entered into a compound library agreement with Pfizer, which expires October 2003, to provide non-exclusive access on a per-compound fee basis to compounds in our Lead Generation Libraries for Pfizer's internal lead generation efforts. Pfizer has the option to gain exclusive rights to compounds upon payment of either a one-time activation fee or annual fees. We retain all ownership of the intellectual property rights to the compounds and to our Lead Generation Libraries as well as any inventions made by our scientists working under this agreement. This agreement is terminable only upon breach or insolvency of a party.

        Takeda Chemical Industries, Ltd.     In July 2001, we entered into a lead generation collaboration agreement with Takeda to create a series of small molecule drug leads against a proprietary Takeda target. Takeda will pay us fees based upon the number of our scientists working on the research phase of the agreement. We are entitled to receive success payments based upon the attainment of certain development milestones and royalties based upon sales of products resulting from the collaboration.

        Tularik Inc.     In February 1999, Tularik acquired a small subset of our Lead Generation Libraries to evaluate the quality of our libraries. Within three months, Tularik initiated a one-year subscription to all of our Lead Generation Libraries. Six months later, Tularik exercised an option to subscribe to our

21

second-year Lead Generation Libraries. We have also expanded our relationship with Tularik by creating focused libraries for a class of targets called orphan nuclear receptors.

        Vertex Pharmaceuticals Incorporated.     In March 2000, Vertex purchased building blocks from our Optimer collection on a non-exclusive basis. This initial introduction led, in August 2001, to a collaboration agreement to discover and develop small molecule drugs directed at two specific targets in the phosphatase protein family. Under this agreement, Vertex provided us with an up-front fee and will provide research funding over three years. We are responsible for the initial drug discovery, including lead generation and lead optimization. Vertex will be responsible for all aspects of clinical development and commercialization, and we are entitled to receive clinical milestone payments. If products are commercialized as a result of this collaboration, we are entitled to additional milestone payments. These milestones, if earned, would be paid on an annual basis for a defined term and are tied to predetermined sales levels.

        Compound library agreements.     We have entered into agreements with customers, including Tularik in June 1999, which Tularik extended in January 2000, and which expires in January 2003; DuPont in August 2000, which expires in December 2005; Hoffmann-La Roche Inc. in June 2001, which expires in June 2006; and Pfizer in October 2001, which expires October 2003, to provide non-exclusive access on a per-compound basis to compounds in our Lead Generation Libraries for their internal lead generation efforts. These customers have the option to gain exclusive rights to compounds upon payment of either a one-time activation fee or annual fees. We retain all ownership of the intellectual property rights to the compounds and to our Lead Generation Libraries as well as any inventions made by our scientists working under these agreements. These agreements are terminable only upon breach or insolvency of a party.

Business Development

        To date, our business development activities have been conducted primarily through direct customer contact by our senior management and scientists and through customer referrals. Because our collaborators are primarily skilled scientists, we use our scientific expertise to initiate and build strong customer relationships. We have relied upon the services of a consulting company, Transpect, Inc., to aid in our business development efforts in Japan. We market our Optimer building blocks through multiple channels, including targeted mailing of a hard copy catalog and through an Internet catalog. We plan to continue to grow our business development resources.

Research and Development

        Our research and development expenses were $4.0 million in fiscal year 2000, $8.3 million in fiscal year 2001 and $13.7 million in fiscal year 2002. We conduct research and development in the following areas:

         Assay development and high throughput screening automation. We are investing in the development of new assay and high-speed screening technologies to more effectively evaluate potential drug compounds for their therapeutic value, including specificity and metabolism, and to increase our screening capacity.

         Informatics. We are continuing our development of database technology to more effectively capture, organize and link the data generated by our scientists and to make this information more seamlessly accessible for any of our drug discovery efforts. In addition, we continue the development of internal software technologies designed to increase the speed and efficacy of our lead generation and lead optimization chemistry.

         Libraries. We have ongoing projects to develop and refine technologies necessary to create high-quality compound libraries composed of drug-relevant compounds that can be rapidly

22

optimized. Our research is focused in the areas of designing drug-relevant building blocks and scaffolds, maximizing drug-like characteristics of our library compounds, optimizing library synthesis processes and maximizing biologically-relevant compound diversity.

         Internal drug discovery projects. We will continue to invest in internal drug discovery programs intended to create a pipeline of proprietary drug candidates. We intend to co-develop and commercialize any resulting drug candidates through collaborations with pharmaceutical and biotechnology companies.

Competitors

        Competition across the range of our drug discovery focus is currently fragmented. We compete with a number of companies in each of the functional areas of drug discovery that we serve. Our major competitors among medicinal chemistry outsourcing companies include: Albany Molecular Research Inc.; ArQule, Inc.; Discovery Partners International, Inc.; and Evotec OAI. Our major competitors among drug discovery companies include: 3-Dimensional Pharmaceuticals, Inc.; Gilead Sciences, Inc.; Tularik Inc.; and Vertex Pharmaceuticals Incorporated. In addition, we compete with the internal research departments of biotechnology and pharmaceutical companies. Many of these companies, some of which are our collaborators and some of which represent market opportunities for us, are developing or already possess internally the technologies we offer. Academic institutions and other research organizations are also conducting research in areas in which we provide our capabilities, either on their own or through collaborative efforts.

Government Regulation

        In the course of our business, we handle, store and dispose of chemicals and biological samples. We are subject to various federal, state and local laws and regulations relating to the use, manufacture, storage, handling and disposal of hazardous materials and waste products. These environmental laws generally impose liability regardless of the negligence or fault of a party and may expose us to liability for the conduct of, or conditions caused by, others. We have not incurred, and do not expect to incur, material costs to comply with these laws and regulations.

        Our customers and collaborators are subject to substantial regulation by governmental agencies in the United States and other countries. Virtually all pharmaceutical products are subject to rigorous preclinical and clinical testing and other approval procedures by the FDA and by foreign regulatory agencies. Various federal and state laws and regulations also govern or influence the manufacturing, safety, labeling, storage, record keeping and marketing of these pharmaceutical products. This approval process is time-consuming and expensive, and there are no assurances that approval will be granted on a timely basis, or at all. Even if regulatory approvals are granted, a marketed product is subject to continual review.

        All facilities and manufacturing processes used in the production of Active Pharmaceutical Ingredients for clinical use in the United States must be operated in conformity with current good manufacturing practices (cGMP) as established by the FDA. We have completed construction of a cGMP manufacturing facility, which will allow us to produce cGMP compliant compounds for Phase I clinical testing. We are validating this capability for compliance with FDA regulations and anticipate being able to initiate our first cGMP manufacturing campaign in the second half of calendar 2002. At that time, our cGMP facility will be subject to periodic regulatory inspections to ensure compliance with cGMP requirements. We could also be required to comply with specific requirements or specifications of our collaborators, which may be more stringent than regulatory requirements. If we fail to comply with applicable regulations, the FDA could require us to cease ongoing research or disqualify the data submitted to regulatory authorities. A finding that we had materially violated cGMP requirements could result in additional regulatory sanctions and, in severe cases, could result in a

23

mandated closing of our cGMP facility, which would materially and adversely affect our business, financial condition and results of operations.

        We are subject to other regulations, including regulations under the Occupational Safety and Health Act, regulations promulgated by the United States Department of Agriculture, and regulations under other federal, state and local laws.

Intellectual Property

        Our success will depend in part on our ability to protect our proprietary software, potential drug candidates and other intellectual property rights. To establish and protect our proprietary technologies and products, we rely on a combination of patent, copyright, trademark and trade secret laws, as well as confidentiality provisions in our contracts with collaborators.

        We attempt to protect our trade secrets by entering into confidentiality agreements with third parties, employees and consultants. Our employees also sign agreements requiring that they assign to us their interests in inventions, original expressions and any corresponding patents and copyrights arising from their work for us. However, it is possible that these agreements may be breached, invalidated or rendered unenforceable, and if so, we may not have an adequate remedy available. Despite the measures we have taken to protect our intellectual property, parties to our agreements may breach the confidentiality provisions or infringe or misappropriate our patents, copyrights, trademarks, trade secrets and other proprietary rights. In addition, third parties may independently discover or invent competing technologies or reverse-engineer our trade secrets or other technology.

        We have also implemented a patent strategy designed to protect technology, inventions and improvements to inventions that are commercially important to our business. As of September 2002, we have four issued United States patents and nine patent applications on file in the United States, including one patent application that has been allowed. We are also pursuing limited patent coverage in foreign countries. Six of our United States patent applications relate to proprietary compounds, including pharmaceutical candidates, and three relate to inventions based on and used in our research efforts. Two of our issued United States patents relating to proprietary pharmaceutical candidates, along with related foreign patent rights, were assigned to us by Amgen in November 1998.

        United States patents issued from applications filed on or after June 8, 1995, have a term of 20 years from the application filing date or earlier claimed priority. All of our patent applications were filed after June 8, 1995. Patents in most other countries have a term of 20 years from the date of filing of the patent application. Because the time from filing patent applications to issuance of patents is often several years, this process may result in a period of patent protection significantly shorter than 20 years, which may adversely affect our ability to exclude competitors from our markets. Our success will depend in part upon our ability to develop proprietary products and technologies and to obtain patent coverage for these products and technologies. We intend to continue to file patent applications covering newly developed products and technologies. We may not, however, commercialize the technology underlying any or all of our existing or future patent applications.

        Patents provide some degree of protection for our proprietary technology. However, the pursuit and assertion of patent rights, particularly in areas like pharmaceuticals and biotechnology, involve complex legal and factual determinations and, therefore, are characterized by some uncertainty. In addition, the laws governing patentability and the scope of patent coverage continue to evolve, particularly in biotechnology. As a result, patents may not issue from any of our patent applications or from applications licensed to us. The scope of any of our patents, if issued, may not be sufficiently broad to offer meaningful protection. In addition, our patents or patents licensed to us, if they are issued, may be successfully challenged, invalidated, circumvented or rendered unenforceable so that our patent rights might not create an effective competitive barrier. Moreover, the laws of some foreign countries may not protect our proprietary rights to the same extent as do the laws of the United States.

24

Any patents issued to us or our strategic partners may not provide a legal basis for establishing an exclusive market for our products or provide us with any competitive advantages. Moreover, the patents held by others may adversely affect our ability to do business or to continue to use our technologies freely. In view of these factors, our intellectual property positions bear some degree of uncertainty.

        The source code for our proprietary software programs is protected both as a trade secret and as a copyrighted work.

Employees

        As of June 30, 2002, we had 245 full-time employees, including 179 scientists, 110 of whom have Ph.D.s and 85 have experience at large pharmaceutical or biotechnology companies. None of our employees are covered by collective bargaining agreements, and we consider our employee relations to be good.