The Japan Advanced Trauma Evaluation and Care (JATEC) education program was introduced in 2002. To examine its effect on the survival of trauma patients, we investigated changes in trauma mortality in Japan in the years after JATEC was introduced. We included patients registered in the Japan Trauma Databank (JTDB) from 2004 to 2011 with clear in-hospital mortality and sufficient data to estimate the Trauma Injury Severity Score (TRISS). Patients were grouped into the early (2004-2006), transition (2007-2008), and late (2009-2011) cohorts. We performed logistic regression analyses after adjusting for TRISS to estimate risk of death in the transition and late cohorts compared with the early cohort. Stratified logistic regression analyses showed which characteristics contributed to the changes in mortality. Of 94,664 patients registered in the JTDB, 47,095 were selected. Adjusted mortality was significantly lower in the late cohort (odds ratio = 0.68; 95% CI, 0.61-0.76) than in the early cohort (reference). Stratification analyses demonstrated significant interactions in patients with or without any chest or abdominal surgery (odds ratio = 0.83 vs 0.68; p < 0.001 in the late cohort) and in patients with TRISS probability of survival <0.5 or ≥ 0.5 (odds ratio = 0.71 vs 0.67 for TRISS probability of survival ≥ 0.5; p < 0.001 in the late cohort). Since 2007, mortality has decreased in patients with mild trauma injury; however, mortality was significantly worse for severely injured patients or patients requiring surgical procedures. These findings suggest that definitive trauma care in Japan should be strengthened. Copyright © 2013 American College of Surgeons. Published by Elsevier Inc. All rights reserved.