Background: In the randomized KEYNOTE-024 trial, pembrolizumab demonstrated robust efficacy in patients with PD-L1≥50% metastatic non-small cell lung cancer (mNSCLC) with median overall survival of (OS) of 26.3 months and 13.4 months in the pembrolizumab and chemotherapy groups, respectively (hazard ratio (HR)=0.62). However, it is unclear how well this efficacy translates to effectiveness in routine practice. Thus, we sought to assess the real-world effectiveness and safety of pembrolizumab for patients with previously untreated PD-L1≥50% mNSCLC compared to platinum-based chemotherapy. Methods: This was a population-based, retrospective cohort study of mNSCLC patients receiving first-line platinum-based chemotherapy (historical controls) from April 1, 2013, to January 17, 2018, or first-line pembrolizumab (cases) from January 17, 2018, to March 31, 2021, in Ontario, Canada. Baseline characteristics and outcomes were ascertained from linked administrative databases. The primary outcome was OS, assessed using Kaplan-Meier and Cox proportional hazards regression methods. Secondary safety outcomes included hospitalization, emergency department visits, specialist visits, and adverse events. To adjust for baseline differences, propensity-score matching was used to match the case and control cohorts (1:1). As a sensitivity analysis, we analyzed OS by histology subgroups (squamous versus non-squamous). Results: A total of 2,284 matched patients were included in our propensity score-matched cohort. The median OS in the pembrolizumab group (13.0 months, 95% confidence interval (CI): 11.8-14.6) was significantly longer than the chemotherapy group (9.2 months, 95% CI: 8.0-10.0), with a HR of 0.81 (95% CI: 0.71-0.92). Though pembrolizumab was also associated with improved OS in both patients with squamous (HR=0.77) and non-squamous histology (HR=0.81), pembrolizumab patients reported significantly more adverse events, specialist visits, and higher 1-year cumulative incidence of direct hospitalizations (11.5% vs 6.8%, p=0.0002). Conclusions: The results of this study confirm the survival benefit of pembrolizumab monotherapy for previously untreated mNSCLC patients with PD-L1≥50% in routine practice, though with increased toxicity and diminished benefit in the real world.