This study was a prospective randomized trial of interferon alfa given either with or without indomethacin in patients with malignant melanoma to determine what effect, if any, indomethacin might have on the toxic, immunomodulatory and therapeutic activity of interferon in this disease. Fifty-three patients were stratified according to performance status and randomized to receive interferon alfa-2b 20 million units/m2 intravenously five days per week for four weeks followed by 10 million units/m2 subcutaneously three times per week either with or without indomethacin 25 mg po three times a day. The overall major response rate was 13 percent (3 complete responders and 3 partial responders among 47 evaluable patients) and was equal on both arms. Indomethacin had no effect on natural killer cell activity or induction of 2'-5' oligoadenylate synthetase in peripheral blood mononuclear cells. The mean maximal temperature elevation induced by interferon was significantly reduced (from 102.1 to 100.7, p = 0.017) by indomethacin, but the incidence and severity of interferon-related fatigue, performance status reductions, elevations in liver function tests, myelosuppression, headache, depression, and confusion was no different on either arm of the study. Indomethacin did not reduce the frequency of interferon dose reductions for toxic side effects and did not allow the administration of higher interferon doses. We conclude that indomethacin can reduce the fever associated with interferon therapy in patients with malignant melanoma treated with interferon without a reduction in the therapeutic or immunomodulatory activity of this agent. Because fever is rarely the dose limiting toxicity of interferon, however, indomethacin is of marginal benefit to such patients receiving interferon in the doses outlined in this study.