Mortality rates for prostate cancer are low in Asia but high in the West. One explanation is the high level of soy consumption in Asia. Soy isoflavones reduce prostate tumor growth in many, but not all, animal models. Elevated levels of serum prostate-specific antigen (PSA) are a marker of prostate tumor growth. Our objective was to determine whether 12-month soy isoflavone supplementation would alter serum PSA concentrations in healthy, older men. The parent study was a double-blinded, parallel-arm, randomized trial in which participants were assigned to consume either a soy protein drink providing 83 mg/day isoflavones (+ISO) or a similar drink with isoflavones removed (-ISO). Participants in the parent study were 85% men. Of the 128 men enrolled in the trial, 112 completed. These men were later contacted for consent to allow their stored sera to be analyzed for PSA and 81 men consented. We measured PSA in serum collected at 0 and 12 months using a commercial radioimmunometric assay. Serum PSA concentrations increased in both groups over the 12-month intervention, but the changes were similar: Geometric mean PSA concentration increased 0.5% more in the +ISO group than in the -ISO group (P = 0.94; 95% confidence interval = -17.3 to 22.2). The proportion of participants having a serum PSA velocity greater than 1 ng/ml/year was similar in the +ISO and -ISO groups (17.6% versus 12.8%; P = 0.54). We found no evidence that a 12-month 83 mg/day isoflavone treatment alters serum PSA concentration or velocity in seemingly healthy men aged 50-80 years.