Background: Small cell bladder cancer (SCBC) is a rare aggressive bladder cancer (BC) variant comprising <1% of BC. Recommended front line etoposide and platinum chemotherapy for SCBC is extrapolated from small cell lung cancer (SCLCA) studies. Data describing small cell bladder cancer treatment are predominantly retrospective. Substitution of etoposide for gemcitabine with cis or carboplatin is widely accepted. This extrapolation is based on histologic similarity between the rare SCBC, and more common SCLCA. Recently, several teams have shown DNA mutation patterns are also more consistent between SCBC and SCLCA, compared to SCBC and urothelial bladder cancers, further supporting the divergence in treatment. These patterns which characterize this “neuronal” subgroup in the TCGA dataset, also reinforce this group has the poorest OS outcome vs other subtypes. Patients with SCBC are nearly uniformly excluded from BC trials, and thus the response to checkpoint inhibition and other novel therapies in SCBC is unknown. As SCBCs often respond to, then quickly recur post chemotherapy, novel active therapies are needed. Immune checkpoint blockade with anti-PD(L)1 antibodies is clinically active, leading to objective responses in approximately 15-25% of patients with urothelial cancer, and 33% in biomarker selected patients with SCLCA. In IMpower133, Atezolizumab + carboplatin + etoposide vs chemotherapy alone demonstrated significant OS advantage (12.3 vs 10.3 months) in frontline ES SCLCA, a practice changing milestone. Like urothelial bladder and SCLCA, we believe SCBC may have dramatic response to checkpoint inhibitors, as shown in small case series from our team and others. Methods: Patients with invasive (cT1-cT4) small cell/neuroendocrine carcinoma of the bladder (MIBC), with or without urothelial cancer component, who are eligible for radical cystectomy and platinum chemotherapy and immunotherapy will be enrolled. Patients with N1 disease within the true pelvis are eligible. Atezolizumab 1200 mg IV Day 1 of every 21 day cycle with etoposide and investigator choice cisplatin or carboplatin chemotherapy x 4 neoadjuvant cycles will be delivered. The primary objective is to assess pathologic complete response rate following protocol therapy. Following cystectomy, Atezolizumab maintenance Q 21 days will continue until unacceptable toxicity or loss of clinical benefit for up to 1 year (e.g., 16 cycles). With type I and type II error rates of 5 and 10% respectively, the first stage will consist of 15 patients. If 6 or fewer patients respond, the study will terminate after the first stage. Otherwise, accrual will continue to 34 patients. If 17 or fewer patients respond at the end of the second stage, the study will terminate after the second stage. If 18 or more patients respond, accrual will continue to 63 patients. Secondary endpoints include rates of DFS and OS. Clinical trial information: NCT05312671.