Background: A phase 2 trial showed improved progression-free survival for atezolizumab plus bevacizumab versus sunitinib in patients with metastatic renal cell carcinoma who express programmed death-ligand 1 (PD-L1). Here, we report results of IMmotion151, a phase 3 trial comparing atezolizumab plus bevacizumab versus sunitinib in first-line metastatic renal cell carcinoma. Methods: In this multicentre, open-label, phase 3, randomised controlled trial, patients with a component of clear cell or sarcomatoid histology and who were previously untreated, were recruited from 152 academic medical centres and community oncology practices in 21 countries, mainly in Europe, North America, and the Asia-Pacific region, and were randomly assigned 1:1 to either atezolizumab 1200 mg plus bevacizumab 15 mg/kg intravenously once every 3 weeks or sunitinib 50 mg orally once daily for 4 weeks on, 2 weeks off. A permuted-block randomisation (block size of 4) was applied to obtain a balanced assignment to each treatment group with respect to the stratification factors. Study investigators and participants were not masked to treatment allocation. Patients, investigators, independent radiology committee members, and the sponsor were masked to PD-L1 expression status. Co-primary endpoints were investigator-assessed progression-free survival in the PD-L1 positive population and overall survival in the intention-to-treat (ITT) population. This trial is registered with ClinicalTrials.gov, number NCT02420821. Findings: Of 915 patients enrolled between May 20, 2015, and Oct 12, 2016, 454 were randomly assigned to the atezolizumab plus bevacizumab group and 461 to the sunitinib group. 362 (40%) of 915 patients had PD-L1 positive disease. Median follow-up was 15 months at the primary progression-free survival analysis and 24 months at the overall survival interim analysis. In the PD-L1 positive population, the median progression-free survival was 11·2 months in the atezolizumab plus bevacizumab group versus 7·7 months in the sunitinib group (hazard ratio [HR] 0·74 [95% CI 0·57–0·96]; p=0·0217). In the ITT population, median overall survival had an HR of 0·93 (0·76–1·14) and the results did not cross the significance boundary at the interim analysis. 182 (40%) of 451 patients in the atezolizumab plus bevacizumab group and 240 (54%) of 446 patients in the sunitinib group had treatment-related grade 3–4 adverse events: 24 (5%) in the atezolizumab plus bevacizumab group and 37 (8%) in the sunitinib group had treatment-related all-grade adverse events, which led to treatment-regimen discontinuation. Interpretation: Atezolizumab plus bevacizumab prolonged progression-free survival versus sunitinib in patients with metastatic renal cell carcinoma and showed a favourable safety profile. Longer-term follow-up is necessary to establish whether a survival benefit will emerge. These study results support atezolizumab plus bevacizumab as a first-line treatment option for selected patients with advanced renal cell carcinoma. Funding: F Hoffmann–La Roche Ltd and Genentech Inc.