The high incidence and long latent period of prostate cancer make it an ideal target for chemoprevention. We have evaluated a series of agents for chemopreventive efficacy using a model in which hormone-dependent prostate cancers are induced in the Wistar-Unilever (WU) rat by sequential treatment with antiandrogen (cyproterone acetate), androgen (testosterone propionate), and direct-acting chemical carcinogen (N-methyl-N-nitrosourea), followed by chronic androgen stimulation (testosterone). This regimen reproducibly induces prostate cancers in high incidence, with no gross toxicity and a low incidence of neoplasia in the seminal vesicle and other non-target tissues. Dehydroepiandrosterone (DHEA) and 9-cis-retinoic acid (9-cis-RA) are the most active agents identified to date. DHEA inhibits prostate cancer induction both when chronic administration is begun prior to carcinogen exposure, and when administration is delayed until preneoplastic prostate lesions are present. 9-cis-RA is the most potent inhibitor of prostate carcinogenesis identified; a study to determine the efficacy of delayed administration of 9-cis-RA is in progress. Liarozole fumarate confers modest protection against prostate carcinogenesis, while N-(4-hydroxyphenyl)retinamide (fenretinide), alpha-difluoromethylornithine, oltipraz, DL-alpha-tocopherol acetate (vitamin E), and L-selenomethionine are inactive. Chemoprevention efficacy evaluations in the WU rat will support the identification of agents that merit study for prostate cancer chemoprevention in humans.