BACKGROUND The roles of modified and selective neck dissections in treating patients with clinical metastatic melanoma and the place of adjuvant radiotherapy are unclear. In the elective setting, the efficacy of various selective dissections also requires clarification. METHODS The prospectively documented experience of the senior author (COB) was analyzed. A total of 175 patients had 183 neck dissections and 92 parotidectomies in 6 years. There were 75 therapeutic and 108 elective operations. Modified or selective neck dissections were performed in 58% of patients with clinical neck metastases. Ali but two elective operations were modified or selective dissections. Postoperative radiotherapy was given to 27 dissected necks. Minimum follow-up was 12 months, and 86% of patients were followed up for 2 years or to neck recurrence. RESULTS Nodes were histologically positive in 80 dissections. The cumulative rate of control of metastatic melanoma in the neck was 86% at 5 years. Neck recurrence developed in 14% of radical dissections, 0% of modified, and 23% of selective dissections performed for clinical disease. Neck recurrence occurred after 5% of elective dissections. Recurrence was 7% among irradiated necks compared to 23% in nonirradiated (p-value not significant). The 5-year survival rate was 50%, and this was significantly worsened by increasing node involvement. CONCLUSIONS Modified radical neck dissection is highly effective in controlling metastatic melanoma in selected patients. Selective dissections are less effective and need further study. Adjuvant radiotherapy appears to decrease the risk of neck recurrence. In the elective setting, recurrence is uncommon following the selective neck dissections described.