BACKGROUND: Laboratory and epidemiologic studies suggest that aspirin and nonaspirin nonsteroidal anti-inflammatory drugs (NSAID) reduce the risk of cancer, possibly via inhibition of the cyclooxygenase enzymes. We evaluated the association of aspirin and nonaspirin NSAIDs with subsequent prostate cancer in a prospective study. We also assessed whether use of these drugs influences serum prostate-specific antigen (PSA) concentration. METHODS: Participants were 1,244 male members of the Baltimore Longitudinal Study of Aging. Use of prescription and over-the-counter drugs was collected by questionnaire and interview at multiple study visits. One hundred forty-one prostate cancer cases diagnosed between 1980 and May 2004 were confirmed by medical record review. We used Cox proportional hazards regression to estimate the rate ratio (RR) of prostate cancer updating drug use over time and taking into account age and year. We used generalized estimating equations to calculate age-adjusted geometric mean PSA concentration by aspirin or nonaspirin NSAIDs use among 933 of the men without prostate cancer, for whom 3,749 PSA measurements in archived sera had been done previously. RESULTS: On 46.0% and 21.5% of the visits, current use of aspirin or nonaspirin NSAIDs (mostly ibuprofen) was reported, respectively. The RRs of prostate cancer comparing ever to never use were 0.76 [95% confidence interval (95% CI), 0.54-1.07] for aspirin, 0.79 (95% CI, 0.54-1.16) for nonaspirin NSAIDs, and 0.71 (95% CI, 0.49-1.02) for either medication. The association for ever use of either aspirin or nonaspirin NSAIDs was suggestively more pronounced in men <70 years (RR, 0.54; 95% CI, 0.27-1.03) than in men >/=70 years (RR, 0.78; 95% CI, 0.50-1.22; P(interaction) = 0.73). The RR for current use of either drug was attenuated relative to ever use. Mean PSA concentration did not differ between users and nonusers of either aspirin or nonaspirin NSAIDs (1.01 versus 0.98 ng/mL, P = 0.56). CONCLUSION: In this prospective study, men, in particular younger men, who had ever used aspirin or nonaspirin NSAIDs had a modest nonstatistically significant lower risk of prostate cancer. The modest inverse association was unlikely due to detection bias that might have resulted if anti-inflammatory drugs had influenced serum PSA concentration.