Background: Race-based differences in ICI efficacy for advanced NSCLC have not been studied due to under-representation of patients of minority background in pivotal trials. We systematically explored real-world differences in outcome in our diverse patient population. Methods: This is a retrospective review of clinical outcome of patients with advanced NSCLC treated with single-agent ICI between 2013 and July 2018 at the Winship Cancer Institute of Emory University. We performed univariate and multivariate analyses for overall survival (OS) and progression free survival (PFS) patients according to self-reported race and of OS according to gender and PD-L1 expression levels. Results: We analyzed clinical data from 90 eligible patients: Median age of 68.5 yrs, 51% male, White (W)/Black(B)/Asians(A) made up 62.3%/30.7%/5%; 36.5% had brain metastasis at the time of ICI initiation. The majority (85.9%) had ECOG PS ≤2; ICI was 1^st line in 15 (16.9%), 2^nd line in 59 (66.3%), 3^rd line in 12 (13.5%) and nivolumab was the most commonly used agent (41.1%) followed by atezolizumab (32.2%) and pembrolizumab (26.7%). The median OS for the entire population was not reached (NR) (95%CI: 15.6, NR) while 12-month and 24-month OS rates were 63.8% (52.8%, 72.8%) and 53.1% (40.2%, 64.4%). The median OS, 12-month and 24-month OS rates for W and B respectively, were 23.6 months vs. NR; HR: 1.02 (95%CI: 0.51-2.04), p = 0.9571; 61.8% (47.7%, 73.2%) vs. 59.3% (38.6%, 75.0%) and 46.0% (27.9%, 62.4%) vs. 53.9% (32.8%, 70.9%). The overall response rate was 16.7%; 23.8% vs. 11% for B and W respectively. The median duration of response was comparable at 3.36 months vs. 2.94 months for W and B. The median PFS and 12-month PFS rate for W and B respectively were 5.5 (3.2, 14.8) vs. 3.0 (1.4, 10.7) months, p = 0.1350 and 40.0% (27.1%, 52.5%) vs. 29.6% (14.1%, 47.0%). Conclusions: Real-world analysis of our institutional experience showed no significant racial disparity in advanced NSCLC patients treated with ICI. Larger multi-institutional studies to include other US minority population would make our findings generalizable.