Background The open-label, randomized, active-controlled KEYNOTE-045 study (NCT02256436) showed that second-line pembrolizumab significantly improved overall survival (OS) of patients with advanced/metastatic urothelial cancer (UC) that progressed after first-line platinum-containing chemotherapy, compared with standard chemotherapy (paclitaxel, docetaxel, or vinflunine). Pembrolizumab is approved for patients with bladder cancer in Japan. Patients and methods Analysis was performed in the subgroup of Japanese patients enrolled in the KEYNOTE-045 study. Coprimary end points were OS and progression-free survival (PFS). Objective response rate (ORR) and safety were secondary end points. Results Fifty-two Japanese patients (pembrolizumab, n = 30; chemotherapy, n = 22) were followed up for a median of 26.1 months. Patients who received pembrolizumab compared with chemotherapy had a 19% lower risk for death (hazard ratio [HR] 0.81, 95% CI 0.44-1.50); after adjusting for baseline covariates, the HR for OS was 0.61 (95% CI 0.32-1.15). The 24-month OS rate was higher with pembrolizumab (26.9% vs 14.3%). PFS was 2.0 and 4.9 months for pembrolizumab and chemotherapy, respectively (HR 1.71, 95% CI 0.95-3.08). ORR was similar for pembrolizumab and chemotherapy (20.0% vs 18.2%); durability of response was higher with pembrolizumab: 67% and 33% of patients, respectively, maintained a response for > 12 months. Treatment-related adverse events, including grade 3-5 events, occurred less frequently with pembrolizumab. Conclusions Pembrolizumab provided durable antitumor activity in patients with locally advanced/metastatic UC that progressed after platinum-containing chemotherapy in the overall population and in the Japanese subgroup; safety profile was consistent with that previously observed for pembrolizumab.