Background: In the phase III KEYNOTE-189 study (NCT02578680), pembrolizumab plus pemetrexed and platinum-based chemotherapy (pemetrexedeplatinum) significantly improved overall survival (OS) and progression-free survival (PFS) in patients with previously untreated metastatic nonsquamous non-small-cell lung cancer (NSCLC) versus placebo plus pemetrexedeplatinum. We report updated efficacy outcomes from the protocol-specified final analysis, including outcomes in patients who crossed over to pembrolizumab from pemetrexedeplatinum and in patients who completed 35 cycles (w2 years) of pembrolizumab. Patients and methods: Eligible patients were randomized 2 : 1 to receive pembrolizumab 200 mg (n ¼ 410) or placebo (n ¼ 206) every 3 weeks (for up to 35 cycles, w2 years) plus four cycles of pemetrexed (500 mg/m 2 ) and investigators' choice of cisplatin (75 mg/m 2 ) or carboplatin (area under the curve 5 mg$min/ml) every 3 weeks, followed by pemetrexed until progression. Patients assigned to placebo plus pemetrexedeplatinum could cross over to pembrolizumab upon progression if eligibility criteria were met. The primary endpoints were OS and PFS. Results: After a median follow-up of 31.0 months, pembrolizumab plus pemetrexedeplatinum continued to improve OS [hazard ratio (HR), 0.56; 95% confidence interval (CI), 0.46-0.69] and PFS (HR, 0.49; 95% CI, 0.41-0.59) over placebo plus pemetrexedeplatinum regardless of programmed death-ligand 1 expression. Objective response rate (ORR) (48.3% versus 19.9%) and time to second/subsequent tumor progression on next-line treatment (PFS2; HR, 0.50; 95% CI, 0.41-0.61) were improved in patients who received pembrolizumab plus pemetrexedeplatinum. Eighty-four patients (40.8%) from the placebo plus pemetrexedeplatinum group crossed over to pembrolizumab on-study. Grade 3-5 adverse events occurred in 72.1% of patients receiving pembrolizumab plus pemetrexedeplatinum and 66.8% of patients receiving placebo plus pemetrexedeplatinum. Fifty-six patients completed 35 cycles (w2 years) of pembrolizumab; ORR was 85.7% and 53 (94.6%) were alive at data cut-off. Conclusions: Pembrolizumab plus pemetrexedeplatinum continued to show improved efficacy outcomes compared with placebo plus pemetrexedeplatinum, with manageable toxicity. These findings support first-line pembrolizumab plus pemetrexedeplatinum in patients with previously untreated metastatic nonsquamous NSCLC.