Background: Molecular alterations (MA) found in brain (Br) mets of NSCLC pts can differ from primary and/or other met sites, which may explain why therapies targeting primary tumors are less effective at preventing and treating intracranial disease. We analyzed the frequency of known driver genes in adv NSCLC pts and the association with overall survival (OS). Methods: This retrospective observational study identified pts from the Flatiron- Foundation Medicine NSCLC Clinico-Genomic Database who were diagnosed with adv NSCLC from 1 Jan 2011 to 31 Oct 2017 and had tumor tissue analyzed at any time following initial diagnosis via targeted DNA sequencing by FoundationOne. Descriptive statistics summarized MA from lung and met sites (Br and non-brain [NB]). OS was measured from adv diagnosis to death or last activity date (censored). Multivariable Cox proportional hazard regression model was used for time-to-event analysis. Results: Of 3257 pts, data were available from lung (n = 1621), Br (n = 180), and NB sites (n = 377): liver (n = 167), bone (n = 124), adrenal (n = 63), and spine (n = 23). Median age at adv diagnosis was 66.2 yrs. TP53(63.3%), KRAS(28.8%), EGFR(15.6%),STK11 (13.5%), and CDKN2A(8.5%) were frequently mutated genes in lung samples. Genes for Br vs NB sites included TP53(70.6%; 64.7%), KRAS(36.1%; 26.5%), EGFR (9.4%; 18.8%), STK11 (18.9%; 12.7%), and CDKN2A(6.1%; 10.1). KEAP1alterations were also present in 10% (Br), 7.4% (NB), and 6% of lung samples. In treated pts, lack of alterations in select genes (STK11, TP53, KEAP1) was associated with longer OS, whereas lack of other alterations (ARID1A, EGFR, ALK, ROS1) was associated with a shorter OS (p < 0.05). Patients with select mutations co-occurring with KRAS had higher risk of death compared to those with KRAS only (p < 0.05). Conclusions: Based on pts with NSCLC whose tumor tissue underwent DNA sequencing, the most frequently altered genes in lung and Br samples included TP53,KRAS, EGFR, STK11,andCDKN2A, with some being significantly associated with OS. Prognosis of NSCLC pts depends on clinical, demographic, and genomic factors and should be carefully considered to optimize clinical outcome.