OBJECTIVE To evaluate the outcome and treatment-related toxicity for patients with axillary lymph node metastases from malignant melanoma treated with surgery and radiation, with or without systemic therapy. METHODS The medical records of 89 consecutive patients with axillary lymph node metastases from malignant melanoma were retrospectively reviewed. All patients underwent axillary dissection and postoperative radiation to a median dose of 30 Gy at 6 Gy/fraction delivered twice weekly. In 3 patients referred with microscopic residual disease, a single boost (4-6 Gy) was given to a reduced field. All but 2 patients were referred because their axillary dissections revealed features believed to predict a 30-50% risk of subsequent axillary recurrence: lymph nodes >/=3 cm in size (54 patients), >/=4 lymph nodes positive (44), the presence of extracapsular extension (69), recurrent disease after initial surgical resection (23), or multiple risk factors (77). Fifty-one patients received systemic therapy before or after radiation therapy. RESULTS At a median follow-up of 63 months, 47 patients had relapsed and 43 patients had died. The actuarial overall and disease-free survival rates at 5 years were 50% and 46%, respectively. The actuarial axillary control and distant metastasis-free survival rates at 5 years were 87% and 49%, respectively. Univariate analysis revealed that the probability of axillary control was inferior when the axillary disease measured >6 cm in size (72% vs. 93%, p = 0.02), the location of the primary tumor was unknown (74% vs. 93%, p = 0.02), the axillary failure occurred within 18 months from diagnosis of the primary melanoma (84% vs. 100%, p = 0.04), or the Breslow thickness was >4 mm (80% vs. 96%, p = 0.04). Additionally, there was an inferior distant metastasis-free and disease-free survival when there were >2 nodes positive for metastatic disease, the primary lesion had a Breslow thickness >4 mm, or the axillary failure occurred within 18 months from diagnosis of the primary melanoma. On multivariate analysis, the significantly inferior distant metastasis-free and disease-free survival seen when >2 nodes were positive or the recurrence occurred within 18 months remained significant. The small number of axillary failures precluded multivariate analysis for axillary control; however, stratified analysis suggested that size >6 cm was the factor most closely associated with subsequent axillary failure. Twenty-six patients developed treatment-related arm edema. Classification according to the severity of edema yielded 5-year actuarial arm edema rates of 21%, 19%, and 1%, for Grade 1 (transient or asymptomatic), Grade 2 (requiring medical intervention), or Grade 3 (requiring surgical intervention) edema, respectively. CONCLUSIONS Adjuvant radiation therapy using a hypofractionated regimen resulted in an 87% 5-year axillary control rate, superior to the 50-70% local control achieved with surgery alone for lymph node metastases from melanoma when high-risk features are present. Improvements are needed for patients with bulky nodal masses >6 cm in size. Mild-to-moderate arm edema was common, but manageable. The degree to which radiotherapy adds to the risk of arm edema after axillary dissection alone cannot be addressed in the present analysis.