Objectives : To evaluate the distribution of serum prostate-specific antigen (PSA) levels as a function of age in men with no evidence of prostate cancer who visited urological clinics. Methods : Simultaneous measurements of total PSA and PSA-a -1-antichymotrypsin (PSA-ACT) were performed on patients who presented at urological clinics in Japan. After excluding 490 patients because of follow-up biopsy findings indicating prostate cancer, patients' history of prostatic surgery and medication affecting the serum PSA level, 1520 patients with PSA levels of less than 20.0 ng/ mL were available for the study.Results : Medians (95th percentile) of the total PSA levels were 0.9 (4.7), 1.2 (5.6), 1.7 (11.0), 2.1 (9.8) and 2.8 (11.0) ng/mL in men in their 40s ( n = 37), 50s ( n = 211), 60s ( n = 488), 70s ( n = 609) and 80s ( n = 175), respectively, whereas those of PSA-ACT were 0.5 (2.9), 0.7 (3.7), 1.1 (7.4), 1.2 (5.9) and 1.6 (6.4) ng/mL, respectively. Both total PSA and PSA-ACT increase with aging, although comparison between the 10-year age groups showed a significant difference in the two molecular forms only between men in their 50s and 60s. Conclusions : The PSA ranges of men who visited urological clinics were higher than those of men participating in prostate cancer screening programs reported in other published studies. An ageassociated increase in PSA similar to screening populations was also observed in urological outpatients. The results of the present study indicate that age-adjusted PSA cut-off levels can be used in outpatient settings, although the PSA reference value derived from the screening population should be carefully applied to symptomatic men of clinical practices.