Background: Rare tumors of the GU tract often exhibit aggressive clinical behavior, yet lack effective standard of care treatment options. SG is an antibody-drug conjugate (ADC) targeting trophoblastic cell surface antigen 2 (Trop2) with a payload of SN-38, a topoisomerase inhibitor. SG monotherapy gained accelerated approval in pts with metastatic urothelial carcinoma (mUC) post platinum and immune-checkpoint inhibitor (ICI) therapy based on the TROPHY-U-01 trial (Tagawa S et al., JCO 2021). Atezolizumab is an anti-PD-L1 ICI approved for use in multiple solid tumors. SMART will evaluate the efficacy and safety of SG monotherapy or SG plus atezolizumab in select rare GU tumors. Methods: SMART is an open-label, non-randomized, Phase 2 trial. Pts must have locally advanced (unresectable) or metastatic GU tumors of the following histologies: small cell carcinoma, squamous cell carcinoma, or primary adenocarcinoma of the bladder or urinary tract, RMC, or squamous cell carcinoma of the penis. Pts may be treatment-naïve or have received any number of prior therapies; pts with small cell carcinoma must have received a platinum-based regimen. Pts treated with a PD-1/PD-L1 ICI will be assigned to Cohort A/Treatment Arm 1 and will receive SG monotherapy (10mg/kg IV, D1 and D8 of 21 day cycles). Pts without prior ICI treatment will be assigned to Cohort B/Treatment Arm 2 and will receive SG (10mg/kg IV, D1 and D8 of 21 day cycles) with concomitant atezolizumab (1200mg IV, D1 of 21 day cycles). Treatment will continue until disease progression (max duration of treatment 5 years) or intolerable side effects. The primary endpoint is objective response rate (ORR) (RECIST v1.1) for each Cohort, with restaging every 3 cycles. Secondary endpoints include safety, clinical benefit rate (CR+PR+SD), median PFS/OS for each Cohort, and ORR for each histology. Exploratory analyses include determination of baseline tumor mutation and transcriptional profile, Trop2 expression, ctDNA/CTCs, peripheral blood immune subsets and cytokine profile correlation with response, and correlation of UGT1A1 allele genotype with response and safety. This study utilizes a Simon minimax two-stage design, with enrollment of up to 43 evaluable pts (accrual ceiling of 60 pts). Cohort A will enroll 11 pts in Stage 1; if 1 or more of the first 11 pts has a response (unacceptably low rate of 5%, p0=0.05; alpha 0.10, beta=0.20), accrual will continue into Stage 2 up to 18 evaluable pts. Cohort B will enroll 16 evaluable pts in Stage 1; if 2 or more pts have a response (unacceptably low rate of 10%, p0=0.10; alpha 0.10, beta=0.10), accrual will continue in Stage 2 up to a total of 25 evaluable pts. This study is open at the NIH Clinical Center (CC); enrollment may open at 1-2 additional centers. Clinical trial information: NCT06161532.