Epidemiologic evidence on the relation between nutrition and prostate cancer is reviewed. Little is known about the etiology of prostate cancer, despite its prominence as the leading cancer among men in the United States. Rational mechanisms for dietary influences on prostate carcinogenesis, including effects on production or metabolism of androgenic hormones, have been proposed, but because few suitable animal models have been developed, the laboratory literature on diet and prostate cancer is sparse. Despite strong ecologic data and largely consistent case-control and cohort data on dietary fat and prostate cancer, the role of this nutrient remains unclear. Few studies, to date, have adjusted the results for caloric intake, and no particular fat component has been consistently implicated. A notable finding is a strong positive association with intake of animal products, especially red meats, but this in itself does not specifically implicate fat. Epidemiologic investigations on vitamin A and carotenoids are divided almost equally between studies showing positive and inverse associations. The evidence from these studies for a protective effect of fruits and vegetables on prostate cancer, unlike many other cancer sites, is not convincing. The data on other dietary components that have been examined with regard to prostate cancer etiology (cadmium, zinc, vitamins C and D, beverages, and legumes) are too incomplete at this time to draw any inferences as to their importance. The evidence for anthropometric associations with prostate cancer is weak. Whereas a clear association with obesity has not been shown, a positive relationship to muscle mass, though not yet established conclusively, further suggests the importance of androgens in this cancer.