To determine the effect of digital rectal examination on the serum prostate specific antigen (PSA) concentration a prospective, randomized, controlled trial involving 143 patients was conducted. Of the patients 86 (60%) had benign prostatic hyperplasia (BPH), 47 (33%) had prostate cancer and 10 (7%) had chronic prostatitis. The study group consisted of 71 men, all of whom had a serum PSA determination followed by a digital rectal examination and then a second serum PSA determination. The control cohort consisted of 72 men, all of whom had 2 serum PSA determinations without an intervening digital rectal examination. The median change in the serum PSA level for the study group was 0.4 ng./ml. compared to -0.1 ng./ml. for the control cohort (p less than 0.0001). For 76% of the study patients the second serum PSA level was greater than the initial value; only 32% of the control patients exhibited a higher second serum PSA level than the initial level (p less than 0.0001). However, only 4 patients with an initial PSA value in the reference range (0.0 to 4.0 ng./ml.) had a post-digital rectal examination value greater than 4.0 ng./ml. and only 1 patient whose presenting serum value was less than 10.0 ng./ml. had a serum PSA level greater than this cutoff point after digital rectal examination. This minimal change in serum PSA after digital rectal examination was independent of the diagnosis (BPH, cancer or chronic prostatitis), initial serum PSA concentration and examiner. Thus, although digital rectal examination had a statistically significant effect on the serum PSA concentration, the clinical significance of a 0.4 ng./ml. median increase appears inconsequential. Based on these findings, physicians should be confident that the serum PSA concentration in the immediate post-digital rectal examination period is accurate and does not compromise clinical use of the tumor marker.