Background: The treatment (tx) landscape for la/mUC has evolved with the use of immunotherapy (IO) for platinum-refractory la/mUC as well as first-line (1L) maintenance therapy (1LM). This cross-sectional survey explored practice patterns for 1L tx/1LM use and clinical decision-making. Methods: Community/academic US oncologists (n = 150) completed an online survey (Sept-Oct 2021) on demographics, 1L tx, 1LM use, attributes in 1L tx selection/1LM use, and factors associated with 1L tx/1LM use. Physicians were dichotomized into 4 pre-specified groups using the median percentage (%) as a cutoff: 1) more frequent 1L prescriber 2) less frequent 1L prescriber (% of pts treated with 1L tx in the past 6 months); 3) more frequent 1LM prescriber 4) less frequent 1LM prescriber (% of pts eligible and received 1LM). Descriptive and bivariate analyses assessing attributes (scored out of 100 points across 16 attributes) in 1L tx selection/1LM use were conducted. Multivariable logistic regression was used to assess factors associated with more/less frequent 1L tx/1LM use. Results: Median time in practice was 15 yrs (range, 2-31; 63% community vs 37% academic setting). The median % of la/mUC pts who received 1L tx was 46% (range, 25-89%). 72 physicians were categorized as more frequent 1L prescribers, while 78 were less frequent 1L prescribers. The median % of pts eligible and received 1LM was 71% (range, 0-100%). 71 physicians were categorized as more frequent 1LM prescribers, while 75 were less frequent 1LM prescribers. Attributes used in 1L tx selection differed among more vs less frequent 1L prescribers: mean scores for efficacy/overall survival (OS), disease control rate (DCR), or rate of grade 3/4 adverse events (AEs) were 23 vs 17, 10 vs 8, and 10 vs 5, respectively (all p < 0.05). Similarly, for more vs less frequent 1LM prescribers, mean scores for efficacy/OS, rate of grade 3/4 immune-mediated AEs, and inclusion in institutional guidelines/pathways were 23 vs 16, 6 vs 4, and 2 vs 4. Oncologists who stated OS, DCR, or rate of grade 3/4 AEs as important factors impacting tx selection were more likely to prescribe 1L tx (all p < 0.05). Regarding 1LM use, oncologists based in the academic setting, those who reported using RECIST 1.1 criteria to assess tx response or agreed 1LM is important to prolong OS were all more likely to prescribe 1LM (all p < 0.05). Those who reported that their institutional guidelines/pathways impact tx decisions or cited prior IO use before metastatic diagnosis as reason not to prescribe 1LM were less likely to prescribe 1LM (all p < 0.05). Conclusions: While several factors were found to be associated with offering 1L tx by US oncologists, including impact on OS and practice setting, variability exists in physicians’ attitudes to 1L tx/1LM use. Studies and interventions to explore shared decision-making for optimal 1L tx selection are needed.