Background: Knowledge of large-scale real-world treatment patterns and clinical outcomes in patients (pts) with metastatic urothelial carcinoma (mUC) is limited. We conducted this study to address the lack of knowledge and identify unmet needs in pts with mUC in real-world clinical practice. Methods: The US nationwide Flatiron Health electronic health records–derived, de-identified database, comprised of 280 oncology practices across the US, was utilized to conduct a retrospective cohort analysis of pts diagnosed with mUC between Jan 1, 2011, and Aug 31, 2020. Baseline pt characteristics were assessed descriptively, and treatment patterns were classified by cisplatin (CIS) eligibility. Kaplan–Meier methods were used to evaluate overall survival (OS) and progression-free survival (PFS). In a subgroup analysis of cisplatin-ineligible (CIS-inelig) pts, a multivariable model adjusting for baseline covariates was used to assess survival outcomes. Results: Of 8183 pts with mUC, median age was 73.0 years at diagnosis. Primary tumor sites were bladder (78.5%), upper tract (20.6%), and urethra (0.9%). Median (range) follow-up from mUC diagnosis was 9.7 (0.2-116.6) months. Of 5855 (71.6%) pts who received first-line (1L) systemic therapy, 1764 (30.1%) were CIS eligible (CIS-elig), 2293 (39.2%) were CIS-inelig, and 616 (10.5%) did not receive CIS despite qualifying ECOG PS (0–1) and renal function; CIS eligibility was unknown in 1182 (20.2%). Among all 1L pts, 4380 (74.8%) received chemotherapy and 1410 (24.1%) received immunotherapy (IO); of the IO users, 1345 (95.4%) received monotherapy. Among CIS-elig pts, CIS plus gemcitabine (GC) and methotrexate, vinblastine, doxorubicin, CIS (MVAC) accounted for 1562 (88.5%) of 1L therapy. Among CIS-inelig pts, carboplatin plus gemcitabine (GCa; 36.1%), pembrolizumab (pembro) monotherapy (18.5%), and atezolizumab (atezo) monotherapy (15.1%) were the most common 1L therapies. Across CIS eligibility groups, the most common second-line therapies included pembro, atezo, and GCa; the most common third-line therapies included atezo, pemetrexed, paclitaxel, and GCa. Median OS (95% CI) was longer in pts who received ≥ 1 line of systemic therapy (14.5 [14.0–15.2] months) than in those who did not receive therapy (6.8 [6.2–7.3] months). Median (95% CI) OS and PFS were also longer in CIS-elig pts (OS, 19.7 [18.2–21.4] months; PFS, 11.5 [10.8–12.1] months) than in CIS-inelig pts (OS, 11.4 [10.8–12.0] months; PFS, 7.0 [6.7–7.4] months), irrespective of receiving treatment. In a subgroup analysis of CIS-inelig pts, 1L IO monotherapy was associated with worse OS than 1L chemotherapy (HR, 1.26; 95% CI, 1.13–1.40, P< 0.0001). Conclusions: This study of > 8000 mUC pts, of whom almost 30% never received systemic therapy, demonstrates real-world treatment patterns in mUC and highlights the substantial unmet need in this population, in particular for CIS-inelig pts.