BACKGROUND: Uncertainty exists about the optimal schedule of adjuvant treatment of breast cancer with aromatase inhibitors and, to our knowledge, no trial has directly compared the three aromatase inhibitors anastrozole, exemestane, and letrozole. We investigated the schedule and type of aromatase inhibitors to be used as adjuvant treatment for hormone receptor-positive early breast cancer. METHODS: FATA-GIM3 is a multicentre, open-label, randomised, phase 3 trial of six different treatments in postmenopausal women with hormone receptor-positive early breast cancer. Eligible patients had histologically confirmed invasive hormone receptor-positive breast cancer that had been completely removed by surgery, any pathological tumour size, and axillary nodal status. Key exclusion criteria were hormone replacement therapy, recurrent or metastatic disease, previous treatment with tamoxifen, and another malignancy in the previous 10 years. Patients were randomly assigned in an equal ratio to one of six treatment groups: oral anastrozole (1 mg per day), exemestane (25 mg per day), or letrozole (2.5 mg per day) tablets upfront for 5 years (upfront strategy) or oral tamoxifen (20 mg per day) for 2 years followed by oral administration of one of the three aromatase inhibitors for 3 years (switch strategy). Randomisation was done by a computerised minimisation procedure stratified for oestrogen receptor, progesterone receptor, and HER2 status; previous chemotherapy; and pathological nodal status. Neither the patients nor the physicians were masked to treatment allocation. The primary endpoint was disease-free survival. The minimum cutoff to declare superiority of the upfront strategy over the switch strategy was assumed to be a 2% difference in disease-free survival at 5 years. Primary efficacy analyses were done by intention to treat; safety analyses included all patients for whom at least one safety case report form had been completed. Follow-up is ongoing. This trial is registered with the European Clinical Trials Database, number 2006-004018-42, and ClinicalTrials.gov, number NCT00541086. FINDINGS: Between March 9, 2007, and July 31, 2012, 3697 patients were enrolled into the study. After a median follow-up of 60 months (IQR 46-72), 401 disease-free survival events were reported, including 211 (11%) of 1850 patients allocated to the switch strategy and 190 (10%) of 1847 patients allocated to upfront treatment. 5-year disease-free survival was 88.5% (95% CI 86.7-90.0) with the switch strategy and 89.8% (88.2-91.2) with upfront treatment (hazard ratio 0.89, 95% CI 0.73-1.08; p=0.23). 5-year disease-free survival was 90.0% (95% CI 87.9-91.7) with anastrozole (124 events), 88.0% (85.8-89.9) with exemestane (148 events), and 89.4% (87.3 to 91.1) with letrozole (129 events; p=0.24). No unexpected serious adverse reactions or treatment-related deaths occurred. Musculoskeletal side-effects were the most frequent grade 3-4 events, reported in 130 (7%) of 1761 patients who received the switch strategy and 128 (7%) of 1766 patients who received upfront treatment. Grade 1 musculoskeletal events were more frequent with the upfront schedule than with the switch schedule (924 [52%] of 1766 patients vs 745 [42%] of 1761 patients). All other grade 3-4 adverse events occurred in less than 2% of patients in either group. INTERPRETATION: 5 years of treatment with aromatase inhibitors was not superior to 2 years of tamoxifen followed by 3 years of aromatase inhibitors. None of the three aromatase inhibitors was superior to the others in terms of efficacy. Therefore, patient preference, tolerability, and financial constraints should be considered when deciding the optimal treatment approach in this setting. FUNDING: Italian Drug Agency.