To assess the effect of cystoscopy, prostate biopsy, and transurethral resection of the prostate (TURP) on the serum prostate-specific antigen (PSA) concentration, 101 patients were evaluated. For cystoscopic examination, 69 men were randomized in a prospective manner to one of three groups: flexible cystoscopy, rigid cystoscopy, and a control cohort. The median change in serum PSA was 0.1 ng/mL following flexible cystoscopy, 0.05 ng/mL after rigid cystoscopy, and 0.05 ng/mL for the control group, in which two serum PSA determinations were obtained without an intervening cystoscopy. The differences between the three groups were not statistically significant. The effect of prostate biopsy and TURP was examined in 32 men. Prostate biopsy caused an immediate elevation in the serum PSA level, with a median increase of 7.9 ng/mL (p < 0.0001). Similarly, TURP produced an elevation in the serum PSA concentration, with a median change of 5.9 ng/mL (p < 0.001). The median time required for the serum PSA value to return to a stable level after prostate biopsy was fifteen days (range: 5-21 days) for men with prostate cancer and seventeen days (range: 3-30+ days) for men without cancer, and eighteen days (range: 12-30+ days) for men who underwent TURP. These findings indicate that a serum PSA determination after either a flexible or a rigid cystoscopy is accurate and reliable. Both biopsy and TURP cause an immediate increase in the serum PSA level, which usually returns to a stable, baseline level within three weeks. However, because in some patients the serum PSA still remained elevated after four weeks, it is recommended that a serum PSA determination not be obtained for at least six weeks after either a prostate biopsy or TURP.