Background: In the randomized, open-label, phase 3 KEYNOTE-361 (NCT02853305) study, superior efficacy was not demonstrated with first-line pembrolizumab ± chemotherapy versus chemotherapy in patients with advanced UC. Although not formally tested, survival outcomes appeared similar with pembrolizumab monotherapy versus chemotherapy (hazard ratio [HR] 0.92; 95% CI, 0.77-1.11). This post hoc exploratory analysis examined efficacy outcomes in patients with complete response (CR) to pembrolizumab or chemotherapy in KEYNOTE-361. Methods: Patients with advanced UC were randomly assigned 1:1:1 to receive first-line pembrolizumab (200 mg IV every 3 weeks for up to 2 years) ± chemotherapy (1000 mg/m² gemcitabine on day 1 and day 8 + cisplatin [70 mg/m²] or carboplatin [area under the concentration curve of 5 mg/ml/min] on day 1 of each 3-week cycle) or chemotherapy alone. Patients were stratified by PD-L1 combined positive score (≥10 vs <10) and investigator’s choice of platinum (cisplatin vs carboplatin). End points of this post hoc exploratory analysis were duration of response (DOR) and progression-free survival (PFS) by RECIST v1.1 by blinded independent central review and overall survival (OS) in patients who achieved a CR in the pembrolizumab monotherapy or chemotherapy arms. Results: Overall, 34 of 307 patients (11.1%) in the pembrolizumab arm and 43 of 352 patients (12.2%) in the chemotherapy arm had a CR. Median follow-up (time from randomization to the data cutoff date) for patients with CR was 30.7 months (range, 22.6-42.3). Median DOR was not reached (NR; range, 4.4+ to 36.1+ months) in the pembrolizumab arm and 12.8 months (range, 2.1+ to 36.3+) in the chemotherapy arm; 75.5% and 37.1% of patients, respectively, remained in CR for ≥24 months. Median PFS was NR (95% CI, 30.3-NR) with pembrolizumab and 15.1 months (95% CI, 8.8-NR) with chemotherapy (HR, 0.32 [95% CI, 0.15-0.70]). The estimated 24-month PFS rates were 75.5% and 42.2% in the pembrolizumab and chemotherapy arms, respectively. Median OS was NR (95% CI, NR-NR) with pembrolizumab and NR (95% CI, 25.1-NR) with chemotherapy (HR, 0.20 [95% CI, 0.06-0.70]). The estimated 24-month OS rates were 94.1% and 69.5% in the pembrolizumab and chemotherapy arms, respectively. Overall, 9 patients (26.5%) in the pembrolizumab arm and 21 patients (48.8%) in the chemotherapy arm experienced disease progression after an initial CR. Among patients who achieved CR, 2 patients (5.9%) in the pembrolizumab arm and 22 patients (51.2%) in the chemotherapy arm received subsequent therapy. Conclusions: This post hoc exploratory analysis suggests CRs with pembrolizumab were more durable than CRs with chemotherapy in patients with advanced UC, with corresponding longer PFS and OS. Further investigation into patient selection is needed to determine which patients could benefit most from first-line pembrolizumab. Clinical trial information: NCT02853305.