8036 Background: In NSCLC, overall frequencies and subtype distribution of EGFR and KRAS mutations differ significantly between Japanese and US patients. In many NSCLC clinical trials, Japanese patient outcomes appear superior to those in western populations. Since 2009, SWOG and Japanese investigators have used the “Common Arm” analytic method to interpret globally-conducted lung cancer clinical trials, providing a mechanism for direct patient-level comparisons in matching trials performed in the US and Japan. S0424 and the Japanese Molecular Epidemiology Study (JME) are large prospective molecular epidemiology studies conducted by SWOG and Japanese investigators and prospectively planned for “Common Arm” analysis. We investigated the effects of mutation distribution on overall survival in early stage NSCLC using data from JME and S0424. Methods: Both studies were designed to accrue representative proportions of ever-smokers (ES) versus never-smokers (NS) in male and female cohorts. Accrual to S0424 was completed in 2011 with 981 pts enrolled (Cheng et al JNCI 2018); JME was completed in 2012 with 957 pts (Kawaguchi et al JCO 2016). A comprehensive questionnaire was used to capture lifestyle, carcinogenic exposures, demographic and clinical data. Five years of follow-up was completed for each study. The association between baseline characteristics and mutation types were analyzed using multivariate logistic regression. The product-limit method was used to estimate overall survival and Cox regression models were fit to estimate hazard ratios. p-values were two-sided and p-values < 0.05 were considered significant. Results: Evaluable pts: 876 for JME; 957 for S0424. Demographic distribution: S0424: 87% Caucasian, 5% Asian, 4% Black; JME: 100% Japanese. As expected, multivariate logistic models adjusting for age, smoking status, gender and BMI showed JME were more likely to have EGFR mutations (p < 0.0001) and less likely to have KRAS mutations (p = 0.02). EGFR mutations were associated with never-smoking (p < 0.0001) and female sex (p < 0.0001); KRAS was associated with smoking (p < 0.0001). Adjusting for age, smoking status and gender, overall survival was greater in JME patients (p < 0.0001,HR = 0.49 (95% CI: 0.40-0.60)). In the subset of pts without an EGFR mutation, survival still favored the JME cohort (p < 0.001, HR = 0.68 (0.53-0.86). Conclusions: Patient level comparison of US (SWOG) and Japanese (JME) patients with early stage NSCLC demonstrated longer survival in Japanese patients even after adjusting for EGFR mutation frequency. Identification of factors that contribute to this finding are under investigation. Female sex remained a risk factor for EGFR mutation-driven disease in both populations. This analysis forms the basis for applying this population-related model in other clinical settings.