Background: The introduction of r-Interferon alpha (rlfn α) into cytostatic chemotherapy of metastatic melanoma appears to improve objective remission rates. Favorable results of early clinical and single institution trials, however, generally could not be confirmed by prospective multicenter studies. With the introduction of fotemustine (FM), a chloro-nitrosurea (CNU), chemically characterized by the graft of an aminophosphonic acid on the CNU-radical, a new active compound became available for the treatment of patients with melanoma, yielding an objective remission rate of 12% (95% CI 6-20%; EORTC1995). Methods: Hoping for an improvement of this modest response rate, a controlled multicenter phase I/II study was started, combining rlfn α2b (10 MU s.c. day 1-28) with FM, initially for 2-, thereafter for 3-weekly doses of 100 mg/m² at days 14, 21, and 28, followed by a 5-week rest period. Results: Out of 31 patients 28 were eligible and evaluable for safety, 26 evaluable for efficacy. We observed 1 complete and 1 partial remission (7.7%, 95% CI 0.9-25.1%), no change occurred in 6 out of 26 (23%). In 24 patients treatment was terminated because of tumor progression. Toxicity was well within acceptable limits and similar to FM alone: Up to 10% WHO grade III thrombocytopenia, 20% grade III and IV granulocytopenia, and moderate gastrointestinal tract toxicity with 10% grade III and 3% grade IV nausea and vomiting, despite conventional antiemetic medication. Conclusions: rlfn α2b in the schedule and dose tested does not seem to improve remission rates of FM alone and may be disadvantageous both in terms of tumor response and quality of life.