Background: Cisplatin-based neoadjuvant chemotherapy is standard of care in MIBC with improved pathologic response and overall survival (OS) compared to RC alone. Pembrolizumab (pembro) is active in high-risk non-muscle invasive and metastatic bladder cancer and is generally well tolerated. This phase II trial evaluated the safety and efficacy of gemcitabine and split-dose cisplatin (GC) + pembro as neoadjuvant therapy prior to RC (NCT02690558). Methods: Patients with clinical T2-4a N0/X M0 urothelial carcinoma of the bladder eligible for RC were enrolled. Patients received pembro 200mg on day 1 with cisplatin 35mg/m² and gemcitabine 1000mg/m² on days 1 and 8 every 3 weeks for 4 cycles, followed by RC within 4-8 weeks. The first 6 patients received full-dose cisplatin (70mg/m² on day 1) and a lead-in pembro dose; this schedule was discontinued for excess toxicity. Primary endpoint was pathologic downstaging rate ( < pT2) with the null and alternative hypothesis rates = 35% and 55%, respectively. Secondary endpoints were toxicity, pT0 rate, event free survival, and OS. Exploratory objectives include association of response with molecular subtype and post-treatment changes in immune microenvironment (predicted neoantigens, immune gene expression, and T cell receptor repertoire). Results: Between May 2016 and July 2020, 39 patients were enrolled (72% cT2, 23% cT3, 5% cT4a) with a median age of 66 and 82% male. Patients received a median of 4 cycles of therapy. All patients underwent RC except one who declined but is included in intention to treat analysis. Rate of < pT2N0 was 56% (22/39) and pT0N0 rate was 36% (14/39). Most common adverse events (AEs) of any grade were thrombocytopenia (29/39; 74%), anemia (27/39; 69%), neutropenia (26/39; 67%), and hypomagnesemia (26/39; 67%). Most common grade 3/4 AEs were neutropenia (16/39; 41%), thrombocytopenia (13/39; 33%), febrile neutropenia (5/39; 13%), and anemia (4/39; 10%). One patient had new onset type 1 diabetes mellitus with ketoacidosis related to pembrolizumab and no patients required steroids for immune-related AEs. Nine patients (23%) discontinued GC + pembro due to AEs, including 4 of the 6 patients who received full-dose cisplatin with pembro lead-in. Survival data are not yet mature and correlative studies are ongoing. Conclusions: Neoadjuvant GC + pembro was generally safe and met its primary endpoint for improved pathologic downstaging. Correlative analyses are ongoing. Additional investigation of this combination is warranted. Clinical trial information: NCT02690558