Gonadotropin-releasing hormone (GnRH) analogues administered for the treatment of advanced prostatic cancer induce a transient increase in plasma testosterone levels during the first week of treatment, often with a secondary rise in plasma levels of prostatic acid phosphatase and a flareup of disease. To determine whether the antiandrogen nilutamide (Anandron) blocks these effects, we carried out a multicenter, placebo-controlled study of nilutamide in men with prostatic cancer treated with the GnRH analogue buserelin. Thirty-six men with disseminated prostatic cancer and elevated plasma levels of prostatic acid phosphatase were randomly assigned to two groups. Group 1 included 17 men who received buserelin (500 micrograms daily subcutaneously) and nilutamide (300 mg daily by mouth); group 2 included 19 men treated with buserelin and placebo. Symptoms were assessed, and plasma was collected before treatment, daily for 14 days, and on days 18, 22, and 29 after the initiation of treatment. Bone pain appeared or worsened in 5 of the 17 men in group 1 and in 12 of the 19 men in group 2 (P less than 0.05). Acute urinary obstruction occurred in one man in group 2. Despite similar changes in the plasma testosterone levels in both groups, the median concentration of plasma prostatic acid phosphatase decreased almost immediately in group 1, but increased transiently, then decreased on day 14 in group 2. Median levels of prostate-specific antigen decreased immediately in group 1 and decreased on day 8 in group 2. We conclude that nilutamide can prevent the adverse consequences of the buserelin-induced transient rise in plasma testosterone levels in men with advanced prostate cancer treated with a GnRH analogue.