Prostate cancer is a leading cause of cancer-related deaths in many countries. Premalignant lesions and invasive cancer occur more frequently in the prostate than in any organ other than the skin. Yet, the incidence of clinically detected prostate cancer is much lower than the histopathological incidence. The slow growth of prostate cancer and the low incidence of clinically manifest disease in some geographical locations or racial/ethnic groups suggest that prostate cancer can be controlled, perhaps by dietary factors. Vitamin D and retinoids have emerged as leading candidates both to prevent and to treat prostate cancer. Many of the activities of these compounds, established from epidemiological studies, research with cell culture and animal models, and clinical trials, are consistent with tumor suppressor effects. However, retinoids may have additional tumor enhancer properties that balance or negate anti-cancer activity. This perhaps explains the overall lack of protective effects of vitamin A compounds against prostate cancer found in epidemiological studies, and the minimal efficacy of retinoids in clinical trials to treat prostate cancer. While current efforts focus on developing strategies to use vitamin D compounds to control prostate cancer, the possibility exists that prostate cancer cells may become resistant to tumor suppressor effects of vitamin D. Analyses of experimental model systems show that prostate cancer cells become less sensitive to vitamin D through loss of receptors or signaling molecules that mediate vitamin D's actions, or through changes in metabolic enzymes that synthesize or degrade vitamin D compounds. The potential promise of exploiting vitamin D to control prostate cancer is tempered by the possibility that prostate cancer, perhaps even at early stages, may develop mechanisms to escape tumor suppressor activities of vitamin D and/or retinoids.