Background: Patients with LA/mUC experience poor outcomes, with few treatment options historically available. In April 2021, SG (a TROP-2-directed antibody drug conjugate) received accelerated approval for patients with LA/mUC who progressed after platinum-based chemotherapy and immune checkpoint inhibitor therapy, based on the pivotal TROPHY U-01 trial (NCT03547973). We sought to evaluate SG safety in patients with LA/mUC treated in routine clinical practice in this study. Methods: A retrospective cohort of patients aged ≥ 18 years with LA/mUC treated with SG in the United States was evaluated using data from the nationwide Flatiron Health electronic health record–derived de-identified database (from Jan 2011 to Oct 2022). Outcomes included patient characteristics, treatment patterns, adverse events (AEs) of interest incidence, and granulocyte-colony stimulating factor (G-CSF) use. Descriptive statistics were used for data analysis. Results: This study included 86SG-treated patients (male, 70%; median age, 71 [range: 25-85] years; 27% with ECOG ≥ 2; 84% treated in community settings). Most patients (71%) received enfortumab vedotin (EV) in the direct prior line. The majority of patients received SG as monotherapy (94%), with most patients overall (99%) receiving SG as second-line (2L) or later. The most common AEs of interest (any grade, overall) were diarrhea (35%), neutropenia (40%) and nausea/vomiting (21%) (Table). Of the 86 patients, 14 (16%) patients were hospitalized, 8 (9%) discontinued SG treatment likely due to AEs/toxicity, and 3 (3%) died during or within 7 days after the line of treatment. During SG treatment, 52% of patients used G-CSF (primary prophylaxis 24%; secondary prophylaxis 19%; therapeutic use 28%). Only 1 (4.5%) patient who had primary G-CSF prophylaxis developed grade ≥ 3 neutropenia. Conclusions: This study is the largest analysis of SG use in a real-world population with LA/mUC to date; it provides insights into the safety of SG in this setting. Compared with prior data, this real-world population was older, with poorer functionality, and the majority received EV in the direct prior line. Observed AEs were consistent with the known safety profile of SG and indicate that this sequencing is feasible.