A randomized, double-blind trial in patients with disseminated, previously untreated prostate cancer (stage D2) was designed to test the hypothesis that maximal androgen blockade improves the effectiveness of the treatment of prostatic cancer. Six hundred three men received leuprolide in combination with either placebo or flutamide, and were followed for a minimum of 5 years. The 303 patients randomly assigned to receive leuprolide and flutamide had a longer progression-free survival and an increase in the median length of survival compared with the 300 patients receiving leuprolide plus placebo. Differences between the treatments were particularly evident for men with minimal disease and good performance status.