BACKGROUND: Patients with metastatic sarcoma have limited treatment options. Nivolumab and ipilimumab are monoclonal antibodies targeting PD-1 and CTLA-4, respectively. We investigated the activity and safety of nivolumab alone or in combination with ipilimumab in patients with locally advanced, unresectable, or metastatic sarcoma. METHODS: We did a multicentre, open-label, non-comparative, randomised, phase 2 study that enrolled patients aged 18 years or older and had central pathology confirmation of sarcoma with at least one measurable lesion by Response Evaluation Criteria In Solid Tumors (RECIST) 1.1, evidence of metastatic, locally advanced or unresectable disease, an ECOG performance status of 0-1, and received at least one previous line of systemic therapy. Patients were assigned to treatment in an unblinded manner, as this trial was conducted as two independent, non-comparative phase 2 trials. Enrolled patients were assigned (1:1) via a dynamic allocation algorithm to intravenous nivolumab 3 mg/kg every 2 weeks, or nivolumab 3 mg/kg plus ipilimumab 1 mg/kg every 3 weeks for four doses. Thereafter, all patients received nivolumab monotherapy (3 mg/kg) every 2 weeks for up to 2 years. The primary endpoint was the proportion of patients with locally advanced, unresectable or metastatic soft tissue sarcoma achieving a confirmed objective response. Analysis was per protocol. This study is ongoing although enrolment is closed. It is registered with ClinicalTrials.gov, number NCT02500797. FINDINGS: Between Aug 13, 2015, and March 17, 2016, 96 patients from 15 sites in the USA underwent central pathology review for eligibility and 85 eligible patients, including planned over-enrolment, were allocated to receive either nivolumab monotherapy (43 patients) or nivolumab plus ipilimumab (42 patients). The primary endpoint analysis was done according to protocol specifications in the first 76 eligible patients (38 patients per group). The number of confirmed responses was two (5% [92% CI 1-16] of 38 patients) in the nivolumab group and six (16% [7-30] of 38 patients) in the nivolumab plus ipilimumab group. The most common grade 3 or worse adverse events were anaemia (four [10%] patients), decreased lymphocyte count (three [7%]), and dehydration, increased lipase, pain, pleural effusion, respiratory failure, secondary benign neoplasm, and urinary tract obstruction (two [5%] patients each) among the 42 patients in the nivolumab group and anaemia (eight [19%] patients), hypotension (four [10%] patients), and pain and urinary tract infection (three [7%] patients each) among the 42 patients in the nivolumab plus ipilimumab group. Serious treatment-related adverse events occurred in eight (19%) of 42 patients receiving monotherapy and 11 (26%) of 42 patients receiving combination therapy, and included anaemia, anorexia, dehydration, decreased platelet count, diarrhoea, fatigue, fever, increased creatinine, increased alanine aminotransferase, increased aspartate aminotransferase, hyponatraemia, pain, pleural effusion, and pruritus. There were no treatment-related deaths. INTERPRETATION: Nivolumab alone does not warrant further study in an unselected sarcoma population given the limited efficacy. Nivolumab combined with ipilimumab demonstrated promising efficacy in certain sarcoma subtypes, with a manageable safety profile comparable to current available treatment options. The combination therapy met its predefined primary study endpoint; further evaluation of nivolumab plus ipilimumab in a randomised study is warranted. FUNDING: Alliance Clinical Trials in Oncology, National Cancer Institute Cancer Therapy Evaluation Program, Bristol-Myers Squibb, Cycle for Survival.