Background: Benefit of IO (PD1 and PD-L1 inhibitors) for mUC was observed in clinical trials but real-world evidence for benefit and clinical outcomes is limited. Methods: This was a retrospective study of adult pts with mUC who initiated IO regardless of PD-L1 expression in the first- (1L cohort) or second-line (2L cohort) setting between 5/1/2016-1/31/2019 in the US Oncology Network (USON), a network of community oncology practices. Descriptive and Kaplan-Meier analyses to evaluate baseline characteristics, treatment patterns and clinical outcomes were conducted using data from USON’s electronic heath record. Results: Among 393 pts in the 1L cohort, median (range) age at IO initiation was 77 (42, 90+), 74% were male, 69% were White, and 19.1% and 4.1% had ECOG performance status (PS) 2 and 3/4, respectively. Among the 366 pts in the 2L cohort, median (range) age at IO initiation was 70 (29, 90+), 74% were male, 71% were White, and 19.7% and 1.4% had ECOG PS 2 and 3, respectively. Median (range) follow-up durations from IO initiation were 4.2 (0, 34.1; 1L cohort) and 4.1 (0, 31.3; 2L cohort) months (mo), during which time 43.1% (1L cohort) and 44.4% (2L cohort) of pts died. Median overall survival (OS) from IO initiation (95% confidence interval [CI]) was 10.6 (9.7, 13.2) mo for the 1L cohort and 9.4 (7.1, 11.5) mo for the 2L cohort; 1-year survival probabilities (95% CI) were 46.6% (40.1%, 52.8%; 1L cohort) and 43.4% (36.8%, 49.8%; 2L cohort). By the end of the follow-up, 48.1% of 1L pts and 47.8% of 2L pts were alive and did not advance to next line of therapy, and 13.5% of 1L and 13.4% of 2L cohort pts advanced to the next line of therapy. Median (95% CI) treatment durations were 2.6 (2.1, 2.9) and 2.8 (2.2, 3.5) mo for the 1L and 2L cohorts, respectively; 6-mo ongoing treatment probabilities (95% CI) were 26.6% (22.2%, 31.2%; 1L cohort) and 31.4% (26.4%, 36.4%; 2L cohort). Conclusions: OS of pts in the real world receiving 1L and 2L IO appears consistent with clinical trial results, although survival follow-up is limited. A minority of pts received post-IO therapy. Future research should examine influence of pt characteristics and PD-L1 expression on treatment choice and outcomes.