Objective: To detect the therapeutic efficacy of CPA and to compare it with surgical or medical castration in advanced prostate cancer Patients and Methods: Patients from 19 Urology Centers with prostate adenocarcinoma of stages T1-4N+MX or T1-4NXM+ were enrolled. A total of 120 patients were randomized to receive CPA 3X100mg/d (Group 1) versus medical or surgical castration (Group 2). The primary endpoints for this trial were overall and disease-spesific survival. Progression-free survival (PSA progression time) and testosterone decrease rate were assessed as secondary endpoints. Progression-free survival probabilities were calculated by the Kaplan-Meier method and comparison of survival probabilities was performed by the Logrank test. Results: The median PSA values were 42ng/dl in both groups at initiation and decreased to 3.0 and 2.1 respectively in 3 months (p>0.05). Castrate testosterone levels were reached in two groups after 3 months therapy (9% and 6.7% of initial values respectively;p>0.05). The data is immature to assess the survival durations, but in median follow-up of 24 months, no difference in regard to PSA-progression was detected in the two groups (p=0.616). Conclusion: This randomized study of CPA and castration in patients with metastatic prostate cancer has not so far revealed any significant differences in progression-free survival. The initial efficacy and tolerability of monotherapy encourages us to comment that this therapy is safe and acceptable.