Background: Histologic subtypes of bladder cancer are associated with poor prognosis and therapy resistance; understanding of their biology can help identify therapies to improve outcomes. We aimed to describe the genomic alteration (GA) landscape of pure urothelial (UC) and histologic subtypes: plasmacytoid (PC), micropapillary (MP), sarcomatoid (SA), small cell/neuroendocrine (SC), squamous cell differentiation (SQ), adenocarcinoma (AD). Methods: Approximately 2165 patients (pts) were selected from the Tempus multimodal database. Selection criteria included diagnosis of bladder cancer (UC or histologic subtypes: PC, MP, SA, SC, SQ, AD) and sequencing via the Tempus xT assay. Genomic and immunotherapy biomarkers, including mutations, fusions, copy number variants, tumor mutation burden (TMB-high defined as ≥10 mutations/Mb) and MSI status were determined for each subtype and compared using Fisher’s Exact and Kruskal-Wallis tests. Results: Among 2165 identified pts, 1738 (80%) had UC (84% pure and 16% mixed histology), Table shows genomic alterations per histologic subtype. Median age at diagnosis was 70. The overall cohort included male (73%), white (84%), and ever smokers (71%). Of 1197 pts with staging information available, 71% tumors were stage IV. TMB-H and MSI-high status were noted in 17% and 1.1% of the overall cohort, respectively, and were relatively similar across histologies. PD-L1 positive status by IHC was noted in 30% in the overall cohort, and statistically lower in PC (7.7%), SC (10%) and AD (8.8%) vs UC (32%), SA (46%) and SQ (45%) (p<0.001). Compared to UC, FGFR3 was significantly lower in SC (9% vs 1%, p=0.003) and AD (9% vs 1%, p=0.001). We noted FGFR2/3 fusions in 2.6% of the entire cohort and all cases were UC. Compared to UC, ERBB2 was significantly higher in MP (13% vs 45%, p<0.001) and lower in SQ (13% vs 2%, p<0.001). ERBB2 amplification was more common in MP (16%) vs 8% in PC, 6.2% in UC, and 1.6% in SC. Conclusions: Distinct genomic alteration patterns were found among different histologic subtypes of bladder cancer and conventional UC. Assessing the genomic landscape of bladder cancer can help identify potential ‘actionable’ targets and biomarkers, and better inform clinical trial designs. Limitations include lack of clinical data annotation, tumor heterogeneity and retrospective study nature.