In a prospective phase II randomized trial, a dose of 100 mg/m2 iv cisplatin every 3 weeks plus forced hydration with or without mannitol diuresis was tested in patients with previously treated advanced malignant melanoma. A total of 67 patients were evaluated: 33 not given mannitol and 34 in the mannitol arm. Two partial remissions (of 2+ and 6.5 months) were achieved in the no-mannitol arm and one complete response and four partial responses (of 1, 2, 2.5, 5.5, and 8 months) were seen in the mannitol arm. Moderate, severe, and life-threatening renal toxicity was less in the mannitol arm, and patients tolerated more doses of cisplatin. The renal toxicity occurred mostly after the first dose of chemotherapy and did not seem to be cumulative. Other side effects were comparable in both arms. We concluded that renal toxicity is less severe in patients treated with cisplatin, hydration, and mannitol and that the use of cisplatin alone or in combination with other active agent(s) should be considered for further evaluation in previously untreated patients with malignant melanoma.