BACKGROUND: Insulin-like growth factors (IGF-1 and IGF-2) bind to the IGF-1 receptor (IGF-1R), increasing cell proliferation and survival. Ganitumab is a monoclonal IgG1 antibody that blocks IGF-1R. We tested the efficacy and safety of adding ganitumab to endocrine treatment for patients with hormone-receptor-positive breast cancer. METHODS: We did this phase 2 trial in outpatient clinics and hospitals. We enrolled postmenopausal women with hormone-receptor-positive, locally advanced or metastatic breast cancer previously treated with endocrine treatment. They were randomly assigned (2:1) with a central randomisation schedule to receive intravenous ganitumab 12 mg per kg bodyweight or placebo in combination with open-label intramuscular fulvestrant (500 mg on day 1, then 250 mg on days 15, 29, and every 28 days) or oral exemestane (25 mg once daily) on a 28-day cycle. Patients, investigators, study monitors, and the sponsor staff were masked to treatment allocation. Response was assessed every 8 weeks. The primary endpoint was median progression-free survival in the intention-to-treat population. We analysed overall survival as one of our secondary endpoints. The study is registered at ClinicalTrials.gov, number NCT00626106. FINDINGS: We screened 189 patients and enrolled 156 (106 in the ganitumab group and 50 in the placebo group). Median progression-free survival did not differ significantly between the ganitumab and placebo groups (3·9 months, 80% CI 3·6-5·3 vs 5·7 months, 4·4-7·4; hazard ratio [HR] 1·17, 80% CI 0·91-1·50; p=0·44). However, overall survival was worse in the the ganitumab group than in the placebo group (HR 1·78, 80% CI 1·27-2·50; p=0·025). With the exception of hyperglycaemia, adverse events were generally similar between groups. The most common grade 3 or higher adverse event was neutropenia-reported by six of 106 (6%) patients in the ganitumab group and one of 49 (2%) in the placebo group. Hyperglycaemia was reported by 12 of 106 (11%) patients in the ganitumab group (with six patients having grade 3 or 4 hyperglycaemia) and none of 49 in the placebo group. Serious adverse events were reported by 27 of 106 (25%) patients in the ganitumab group and nine of 49 (18%) patients in the placebo group. INTERPRETATION: Addition of ganitumab to endocrine treatment in women with previously treated hormone-receptor-positive locally advanced or metastatic breast cancer did not improve outcomes. Our results do not support further study of ganitumab in this subgroup of patients. FUNDING: Amgen.