Purpose: To evaluate the efficacy and toxicity of stereotactic radiotherapy in the treatment of brain metastases of melanoma. Patients and Methods: From 1994 to 2001, 25 patients presenting with 61 metastases of cutaneous melanoma were treated with radiosurgery. Median age was 47 years (range: 25–73 years) and median Karnofski performance status 80 (range: 50–100). Twenty patients had one radiosurgery, 5 had two or three. Median metastasis diameter was 21 mm (range: 6–54.4 mm), and median metastasis volume was 1.7 cm3 (range: 0.4–25.6 cm3). Irradiation was delivered by a linear accelerator. Median minimal dose was 14.1 Gy (range: 10–19.4 Gy), and median maximal dose was 20.5 Gy (range: 16–48 Gy). Results: Mean follow-up was 12.6 months (range: 1–85 months). Five metastases progressed (9.8%), 2–12 months after radiosurgery. Three-, 6- and 12-month local control rates were 95 ± 3, 90 ± 5 and 84 ± 7%, respectively. By univariate analysis, only absence of extracranial tumor was a prognostic factor of local control. Three-, 6- and 12-month brain-disease-free survival rates were 75 ± 9, 68 ± 11 and 38 ± 13%, respectively. According to univariate analysis, only the Score Index for Radiosurgery in brain metastases (SIR) was a prognostic factor of brain-event-free survival (p = 0.03). Median survival was 8 months. Three-, 6- and 12-month overall survival rates were 75 ± 9, 53 ± 10, and 29 ± 10%, respectively. According to univariate analysis, extracranial controlled disease status (p = 0.03), and SIR (p = 0.04) were prognostic factors for overall survival. According to multivariate analysis, none was an independent prognosticator for overall survival. Complications were minimal. Conclusion: Radiosurgical treatment of brain metastases of melanoma is effective and accurate. The use of radiosurgery alone is an appropriate management strategy for many patients with brain metastases of melanoma.