Multicentricity, heterogeneity, and sampling error complicate the study of the natural history of prostate cancer. Carcinomas previously termed "latent" are probably similar to those detected clinically. The diagnosis of carcinoma is only made following biopsy and histologic confirmation. Estimates of grade and stage made at the time of detection are prone to sampling error and are likely to change following examination of radical prostatectomy specimens. In regions with lower life expectancies, the problem of prostate cancer becomes a lower priority due to its association with aging. Access to care, numbers of medical facilities, and differing public health priorities make international rates difficult to compare. However, in view of the high worldwide prevalence of the disease, further epidemiologic studies of prostate cancer are warranted.