Address for correspondence: Jonathan B. Tucker, Chemical and Biological Weapons Nonproliferation Project, Center for Nonproliferation Studies, Monterey Institute of International Studies, 425 Van Buren Street, Monterey, CA 93940, USA; fax: 831-647-6534; e-mail: jtucker@miis.edu. Since the Japanese doomsday cult Aum Shinrikyo released sarin nerve gas on the Tokyo subway in March 1995, killing 12 people, terrorist incidents and hoaxes involving toxic or infectious agents have been on the rise. Before the late 1990s, the Federal Bureau of Investigation (FBI) typically investigated a dozen cases per year involving the acquisition or use of chemical, biological, radiologic, or nuclear materials; however, FBI opened 74 such investigations in 1997 and 181 in 1998 (1). Although 80% of these incidents have been hoaxes, some were unsuccessful attacks (2). The vulnerability of civilian populations to chemical, biological, radiologic, or nuclear terrorism has been widely discussed, but information on historical cases is anecdotal and often inaccurate (3). Without a realistic threat assessment based on solid empirical data, government policymakers lack the knowledge they need to design prudent and cost-effective programs for preventing or mitigating future incidents. Responding to this knowledge gap, the Chemical and Biological Weapons Nonproliferation Project at the Monterey Institutes Center for Nonproliferation Studies has compiled an open-source database of all publicly known cases from 1900 to the present in which domestic or international criminals or terrorists sought to acquire or use chemical, biological, radiologic, or nuclear materials. As of January 31, 1999, the database contained 415 incidents, both domestic and international. Each entry draws on multiple sources and includes a detailed description of the event and a list of citations. The project has conducted a preliminary analysis of the data to discern patterns over time in the frequency of such incidents, the underlying motives, and the choice of agent and target. The ultimate goal is to identify which types of individuals or groups are most likely to acquire and use toxic or infectious materials and for what purposes. Since the Monterey Database has been compiled from journalistic accounts and other unclassified sources, it may not be comprehensive or fully accurate. Incidents have been recorded only if they came to the attention of law enforcement or the news media, so the database does not include events that were not detected or whose existence remains secret. Despite these limitations, the information in the database indicates trends and patterns of behavior that may assist intelligence and law-enforcement personnel in focusing their monitoring efforts.