Background: Atezo (anti–PD-L1) inhibits PD-L1 to restore anticancer immunity; bev may enhance atezo efficacy by inhibiting VEGF immunosuppression and promoting T-cell tumor infiltration. IMpower150 is the first randomized Ph 3 trial evaluating atezo + chemo (carboplatin [C] + paclitaxel [P]) ± bev vs CP + bev in 1L NSQ NSCLC. PFS benefit was observed with atezo + CP + bev vs CP + bev regardless of PD-L1 expression. Here we present the IMpower150 interim OS results. Methods: 1202 patients (pts) received atezo 1200 mg + C AUC 6 + P 200 mg/m² (Arm A) or atezo + CP + bev 15 mg/kg (Arm B) vs CP + bev (Arm C) IV q3w for 4 or 6 cycles per investigator (INV); then maintenance atezo, atezo + bev, or bev, respectively. Co-primary endpoints were INV-assessed PFS in the ITT-WT (EGFR/ALK WT) and in WT pts with expression of a tumor T-effector gene signature (Teff-high WT) and OS in the ITT-WT. Data cutoff: 1/22/2018. Results: 349, 359, and 337 ITT-WT pts were enrolled in Arms A, B, and C, respectively, with median age 63 y, 62% male, 85% current/previous smokers, and 42% ECOG PS 0. With 13.5 mo min FU, OS was improved in Arm B vs C (HR, 0.78 [95% CI: 0.64, 0.96]; P = 0.016) in the ITT-WT; populations of interest are shown in the Table. Arm A vs C OS HR was 0.88 (95% CI: 0.72, 1.08; P = 0.204). In all treated patients, Gr 3-4 treatment-related AEs occurred in 43%, 57%, and 49% of pts in Arms A, B, and C, respectively. Conclusions: IMpower150 showed a significant OS benefit with atezo + CP + bev vs CP + bev in 1L NSQ NSCLC, and no new safety signals were seen. Clinical trial information: NCT02366143 IC, tumor-infiltrating immune cells; NE, not estimable; TC, tumor cells. ^a WT excludes pts with EGFR or ALK genomic alterations. ^b Present at baseline. TC1/2/3 or IC1/2/3 = PD-L1+ ≥ 1% of TC or IC; TC0 and IC0 = PD-L1+ < 1% of TC and IC