Serum PSA-based early detection for prostate cancer has been studied fairly extensively for the past several years. It appears that we can state fairly categorically what the relative performances of total serum PSA, DRE, and TRUS are in detecting early-stage prostate cancer; that initial screening is effective in detecting histologically significant and pathologically organ-confined prostate cancer; that annual, serial, repetitive screening, at least over a 4- to 5-year horizon, does not overdetect prostate cancer, and that the results of early detection will improve as our ability to use certain PSA transformations such as PSA density, PSA slope, age-specific PSA adjustment, and knowledge of free versus total serum PSA is better characterized. These advances in our ability to diagnose early-stage prostate cancer likely will be coupled with an increased ability to predict the behavior, curability, and significance of individual tumors. It is hoped that information soon will be available to allow physicians to categorize an individual tumor as insignificant, significant and surgically curable, or significant and incurable by standard approaches. This ability, coupled with the demon strated ability to detect prostate cancer, will make an even more compelling argument for widespread PSA-based screening. At present, annual DRE and total serum PSA measurements are recommended for men older than 50 and among younger men at high risk for prostate cancer. All suspicious DRE findings should be evaluated with prostatic biopsy. Among younger men, PSA levels over 2.5 ng/mL should be considered worrisome and further evaluated. For men older than 65, serum PSA levels above 4 ng/mL should be considered abnormal and warrant biopsy. Men with persistent serum PSA elevation and a negative biopsy should undergo repeat biopsy at least once, and perhaps more often if PSA slope exceeds 0.75 per year, if density is greater than 0.10, or if f-PSA is less than 20%.