BACKGROUND The American Cancer Society-National Prostate Cancer Detection Project is a prospective study of the feasibility of early prostate cancer detection by digital rectal examination (DRE), transrectal ultrasound (TRUS), and prostate specific antigen (PSA). Two thousand nine hundred ninety-nine men not previously suspected of having prostate cancer have been entered on study in ten participating clinical centers. METHODS The study protocol requires these men to undergo testing with the three detection methods under investigation on an annual basis for up to 5 years. Data currently are available on 1972 men who have completed protocol requirements for two sequential series of examinations. RESULTS At the first visit, it was recommended that 16.5% of subjects undergo biopsy based on the TRUS and/or DRE findings. Seventy-three (3.7%) men were found to have cancer after the first series of examinations. When the 1899 men who did not have cancer were reexamined, in 11.4%, a biopsy was recommended, and an additional 33 (1.7%) cancers were detected. Cancer detection rates were higher in older men, and the sensitivity of TRUS was higher than that of DRE at both examinations. The positive predictive values (PPV) of TRUS and DRE varied significantly depending on the PSA level; PPV decreased at the second examination. Positive findings during the second examination were associated with elevated PSA levels during the earlier visit and rising PSA levels between examinations. Analyses of the 106 cancers detected currently showed a preponderance of early-stage cancer and low Gleason scores. CONCLUSIONS It may be possible to increase the early detection of prostate cancer by periodic examination by a combination of detection modalities.