Target populations for chemoprevention trials should include those at higher than average risk for the development of prostate cancer as defined by explicit epidemiologic and genetic criteria. Such populations include a "primary prevention" group without histologic or clinical evidence of cancer, and several clinical models of "secondary prevention," including those with clinically evident disease prior to definitive therapy and those at high risk of recurrence after therapy based on histological or biochemical status. Each risk group and clinical model has potential advantages and disadvantages, and the mechanisms that underlie disease development and progression in each group may be unique. These observations give rise to many potential clinical trials of specific agents. These trials should also include collection of data on potentially confounding influences on disease development and progression.