BACKGROUND: Adjuvant radiotherapy is recommended for patients with melanoma after lymphadenectomy. We previously showed this treatment reduced risk of repeat lymph-node field cancer in patients with a high risk of recurrence but had no effect on overall survival. Here, we aim to update the relapse and survival data from that trial and assess quality of life and toxic effects. METHODS: In the ANZMTG 01.02/TROG 02.01 randomised controlled trial, we enrolled patients who had undergone lymphadenectomy for a palpable lymph-node field relapse and were at high risk of recurrence at 16 hospitals (11 in Australia, three in New Zealand, one in Netherlands, and one in Brazil). We randomly assigned patients (1:1) to adjuvant radiotherapy (48 Gy in 20 fractions, given over a maximum of 30 days) or observation, stratified by institution, areas of lymph-node field (parotid and cervical, axilla, or groin), number of involved nodes (≤3 vs >3), maximum involved node diameter (≤4 cm vs >4 cm), and extent of extracapsular extension (none, limited, or extensive). Participants, those giving treatment, and those assessing outcomes were not masked to treatment allocation, but participants were unaware of each other's treatment allocation. In this follow-up, we assessed outcomes every 3 months from randomisation for the first 2 years, then every 6 months up to 5 years, then annually. The primary endpoint was lymph-node field relapse as a first relapse, assessed in patients without major eligibility infringements (determined by an independent data monitoring committee). We assessed late adverse effects (occurring >90 days after surgery or start of radiotherapy) with standard criteria in the as-treated population. This study is registered with ClinicalTrials.gov, number NCT00287196. FINDINGS: Between March 21, 2003, and Nov 15, 2007, we randomly assigned 123 patients to adjuvant radiotherapy (109 eligible for efficacy assessments) and 127 to observation (108 eligible). The final follow-up date was Nov 15, 2011. Median follow-up was 73 months (IQR 61-91). 23 (21%) relapses occurred in the adjuvant radiotherapy group compared with 39 (36%) in the observation group (adjusted hazard ratio [HR] 0·52 [95% CI 0·31-0·88], p=0·023). Overall survival (HR 1·27 [95% CI 0·89-1·79], p=0·21) and relapse-free survival (0·89 [0·65-1·22], p=0·51) did not differ between groups. Minor, long-term toxic effects from radiotherapy (predominantly pain, and fibrosis of the skin or subcutaneous tissue) were common, and 20 (22%) of 90 patients receiving adjuvant radiotherapy developed grade 3-4 toxic effects. 18 (20%) of 90 patients had grade 3 toxic effects, mainly affecting skin (nine [10%] patients) and subcutaneous tissue (six [7%] patients). Over 5 years, a significant increase in lower limb volumes was noted after adjuvant radiotherapy (mean volume ratio 15·0%) compared with observation (7·7%; difference 7·3% [95% CI 1·5-13·1], p=0·014). No significant differences in upper limb volume were noted between groups. INTERPRETATION: Long-term follow-up supports our previous findings. Adjuvant radiotherapy could be useful for patients for whom lymph-node field control is a major issue, but entry to an adjuvant systemic therapy trial might be a preferable first option. Alternatively, observation, reserving surgery and radiotherapy for a further recurrence, might be an acceptable strategy. FUNDING: National Health and Medical Research Council of Australia, Cancer Council Australia, Melanoma Institute Australia, and the Cancer Council South Australia.