Background: Patients (pts) with mNSCLC with active brain metastases (BM) are often excluded from clinical trials; data on efficacy and safety of immunotherapy in this population are limited. We compared outcomes of pts with mNSCLC with and without BM who received pembrolizumab-based therapy. Methods: We conducted a retrospective single-center study of pts with mNSCLC treated with pembrolizumab (P) with or without chemotherapy. Progression-free survival (PFS) and overall survival (OS) were determined by Kaplan-Meier methodology and compared using multivariable Cox regression and log rank testing. Results: We identified 587 consecutive pts with mNSCLC who began P-based therapy between 8/2013 and 12/2018: 306 (52%) female, median age 67 years (range 32-98), 437 (74%) adenocarcinoma, and 508 (87%) former/current smokers. 388 (66%) patients received P in first line therapy, and 334 (57%) received single-agent P. 131 pts (22%) had detectable BM at baseline (start of P-based therapy). Pts with BM were younger (median 65 y vs 68 y, p < 0.01) and more likely to have adenocarcinoma (86% vs. 71%, p < 0.01) and baseline steroid use (22% vs 1%, p < 0.01). Presence of BM did not differ by race, sex, line of therapy, treatment regimen, or PD-L1 status. Of the 131 patients with detectable BM on pre-treatment brain MRI, 55 (42%) had stable BM as a result of prior local therapy, while 76 (58%) had active (new or growing) BM on pre-treatment imaging. Most patients with active BM underwent radiation therapy (RT) in either the 30 days before (n = 46) or 30 days after (n = 17) P start; of the remaining 13 treated with P-based therapy alone, intracranial responses included 2 CR, 2 PR, 3 SD, and 4 PD. As of 1/1/2020, with 15-month median follow up, there was no difference in mPFS (9.2 vs 7.3 months, p = 0.41) or mOS (18.3 vs 18.0 mo, p = 0.67) between pts with and without BM in our P-treated cohort. On multivariable analysis, female sex, ECOG 0-1, adenocarcinoma histology, and P as first line therapy were associated with improved PFS and OS. Presence of BM, baseline steroid use, and timing of local RT (before vs. after P) were not associated with inferior survival. Conclusions: In our single-center experience of pts with mNSCLC treated with P, pts with and without BM had similar PFS and OS. We observed several intracranial responses to P-based therapy alone, but most pts with active BM underwent local RT. mNSCLC pts with BM should be considered for P-based therapy; BM may be treated with RT immediately before or even after P with similar survival outcomes.