Three hundred and twenty-two postmenopausal patients with primary breast cancer and ipsilateral axillary node involvement were randomized to receive aminoglutethimide and hydrocortisone or placebo for 2 years in a double blind randomized trial between April 1980 and March 1985. Two hundred and eighty-six patients were eligible for the study of whom 145 received active drug and 141 received placebo. At the present time significantly fewer patients have relapsed or died without previous relapse in the treatment arm (P = 0.002); 43 of 145 (30%) patients receiving aminoglutethimide have relapsed or died compared with 63 of 141 (40%) of those receiving placebo. Local recurrence is also significantly reduced (P = 0.002) since only 6 patients receiving active treatment developed local recurrence compared to 21 receiving placebo. Side effects were severe enough to necessitate complete withdrawal or reduction of therapy in 27 of 145 (19%) in the treatment arm of the study compared with 21 of 141 (15%) in the placebo arm. A single treatment-related death occurred, due to agranulocytosis. Aminoglutethimide and hydrocortisone therefore delay relapse after surgery for primary breast cancer in postmenopausal women. It is too early to assess any effect on overall survival.