We undertook a systematic review of 41 randomised studies in disseminated melanoma, identified by a comprehensive search. We aimed to investigate rates of response to various treatment modalities and the outcome for the patients. We analysed seven studies that compared polychemotherapy with single-agent dacarbazine, six that compared different chemotherapeutic schedules with each other, five on the addition of tamoxifen to a reference therapy, and six that included non-specific immunostimulators. In 17 studies, the addition of interferon alfa, interleukin 2, or both, to a reference therapy was investigated, including trials with biochemotherapy. Many trials had small sample sizes and did not report a power analysis; not all were analysed by intention to treat. Although some treatment regimens, especially polychemotherapeutic schedules, seem to increase response rates, none of the treatment schedules was proven to prolong overall survival. Patients with disseminated melanoma should be treated with well-tolerated drug regimens, such as single-agent treatments or in combination with interferon alfa. Systemic treatments should preferably be investigated in randomised trials so that the potential benefits of new treatment concepts can be thoroughly examined.