The incidence of brain cancer is rising in the United States while the causes remain largely unknown. Epidemiologic studies indicate that individuals working in agriculture have an increased risk of brain cancer. The National Institute for Occupational Safety and Health is conducting a case-control study of incident brain cancer cases in Iowa, Michigan, Minnesota, and Wisconsin to evaluate the risk associated with several environmental exposures, in particular agricultural pesticides. Hundreds of different pesticides are used in agriculture and it is not feasible to evaluate the association between brain cancer and exposure to each of these chemicals; therefore, a strategy was developed to identify which pesticides would be targeted in the study. First lists of pesticides were created, documenting usage in each of the four states and the United States as a whole, by using data from reports prepared by the U.S. Department of Agriculture and Departments of Agriculture and land grant colleges within the four states. Then the following factors were considered in prioritizing pesticides for evaluation in the study: total volume of use prior to 1985, ranking of use in the four states and the United States as a whole by pesticide category, and toxicological evidence of carcinogenic, teratogenic, or mutagenic effects. Pesticide usage prior to 1985 was determined to allow for a minimum 10-year latency for the incident brain cancer cases diagnosed in 1995 or later. The selected pesticides include 56 herbicides, 49 insecticides, 12 fungicides, and 17 fumigants, accounting for over 99% of the total pounds of herbicides and insecticides and over 98% of the total pounds of fungicides and fumigants applied pre-1985. Prompt lists of the pesticides are sent to study participants a few days before the study questionnaire is administered to allow them time to recall past use of pesticides; the lists include the common chemical names, trade names, the crops that the pesticides are most commonly used on, and the years that the pesticides have been marketed. The methods used to select this subset of 134 pesticides document historical usage and may be useful in prioritizing pesticides for other research studies.