BACKGROUND Prostate-specific antigen (PSA) is used as an outcome measure for relapsed disease in prostate cancer. Nonetheless, there are considerable concerns about its indiscriminate use as a surrogate endpoint for cell growth or survival. We hypothesized that treatment with a luteinizing hormone releasing hormone (LHRH) analog would decrease PSA levels even in the absence of malignant disease. METHODS We determined testosterone and PSA levels in 30 healthy volunteers after a single intramuscular injection of a LHRH depot formulation. Testosterone and PSA levels were quantified by radioimmunoassay and electrochemi-luminescence immunoassay, respectively. RESULTS After an initial flare-up during the first 3 days testosterone decreased reaching castration levels in 18 of the 30 young men (60%). After the nadir on day 28, testosterone levels increased to normal again. Changes in PSA paralleled those of testosterone. Castration reduced PSA levels by 29% (95% CI 19%-39%) compared to baseline (p<0.0001). CONCLUSIONS LHRH superagonists decrease PSA levels by testosterone deprivation. Conferring these findings to tumor patients, decreases in PSA after treatment with LHRH analogs might not only reflect disease regression but also a direct testosterone mediated effect on PSA. Thus, PSA levels should be cautiously interpreted when patients receive hormonal therapy.