IFN-gamma is a potent immunomodulator, which has activity against melanoma in vitro and in murine models. However, preclinical data suggests that the optimal therapeutic and immunomodulatory dose may not be the maximally tolerated clinical dose. We conducted a Phase II/III trial in good prognosis patients with metastatic melanoma to determine whether a therapeutic and immunomodulatory dose-response curve of IFN-gamma could be identified, and whether the two could be correlated. Ninety-eight patients with metastatic melanoma were randomized to one of seven dose levels of IFN-gamma ranging from 0.01 to 0.90 mg/m2. All patients were required to have s.c., skin, soft tissue, or nodal disease, although visceral disease was also allowed, and no more than one prior chemotherapy regimen. Patients received IFN-gamma as a 1-h i.v. infusion three times per week for at least 8 weeks or until progressive disease. Ninety-five patients were eligible for toxicity evaluation; 81 were eligible for tumor response. Four patients responded to therapy (response rate, 5%) at three dose levels: two patients at 0.01 mg/m2 and one each at 0.5 and 0.9 mg/m2. The duration of response ranged from 5 to 58 weeks. Toxicities were typical of IFNs and included flu-like constitutional symptoms. No dose-response relationship was identified for efficacy. A dose-response relationship for toxicity was observed only for fever and chills (p = 0.035) and hepatic toxicity (p = 0.034). IFN-gamma has minimal activity in metastatic melanoma, and a therapeutic dose-response curve could not be identified. Although potent dose-dependent effects on immunomodulation were identified (J. M. Kirkwood, J. Bryant, J. H. Schiller, M. M. Oken, E. C. Borden, and T. L. Whiteside. Immunomodulatory function of interferon gamma in patients with metastatic melanoma: results of a phase IIB trial in subjects with metastatic melanoma: ECOG Study E4987, submitted for publication), this biological activity does not translate into therapeutic activity in the metastatic disease setting in this trial.