Ipilimumab is a human monoclonal antibody that blocks cytotoxic T-lymphocyte antigen 4 and has shown promising activity in advanced melanoma. We aimed to ascertain the antitumour efficacy of ipilimumab in patients with advanced melanoma. We undertook a randomised, double-blind, phase 2 trial in 66 centres from 12 countries. 217 patients with previously treated stage III (unresectable) or stage IV melanoma were randomly assigned a fixed dose of ipilimumab of either 10 mg/kg (n=73), 3 mg/kg (n=72), or 0.3 mg/kg (n=72) every 3 weeks for four cycles (induction) followed by maintenance therapy every 3 months. Randomisation was done with a permuted block procedure, stratified on the basis of type of previous treatment. The primary endpoint was best overall response rate (the proportion of patients with a complete or partial response, according to modified WHO criteria). Efficacy analyses were done by intention to treat, whereas safety analyses included patients who received at least one dose of ipilimumab. This study is registered with ClinicalTrials.gov, number NCT00289640. The best overall response rate was 11.1% (95% CI 4.9-20.7) for 10 mg/kg, 4.2% (0.9-11.7) for 3 mg/kg, and 0% (0.0-4.9) for 0.3 mg/kg (p=0.0015; trend test). Immune-related adverse events of any grade arose in 50 of 71, 46 of 71, and 19 of 72 patients at doses of 10 mg/kg, 3 mg/kg, and 0.3 mg/kg, respectively; the most common grade 3-4 adverse events were gastrointestinal immune-related events (11 in the 10 mg/kg group, two in the 3 mg/kg group, none in the 0.3 mg/kg group) and diarrhoea (ten in the 10 mg/kg group, one in the 3 mg/kg group, none in the 0.3 mg/kg group). Ipilimumab elicited a dose-dependent effect on efficacy and safety measures in pretreated patients with advanced melanoma, lending support to further studies at a dose of 10 mg/kg. Bristol-Myers Squibb. Copyright 2010 Elsevier Ltd. All rights reserved.