Until now, we routinely administered oxygen to trauma patients in prehospital settings irrespective of whether oxygen delivery affected the prognosis. To determine the necessity of prehospital oxygen administration (POA) to trauma patients, we aimed to assess whether POA contributed to in-hospital mortality.This was a multicenter propensity-matched cohort study involving 172 major emergency hospitals in Japan. During 2004 to 2010, 70,683 patients with trauma aged ≥15 years were eligible for enrolment. The main outcome measures were survival until hospital discharge after POA, and propensity score analyses were used to adjust for patient factors and hospital site.Of 32,225 trauma patients, 19,985 (62.0%) were administered oxygen by the emergency medical services in prehospital settings and 12,240 (38.0%) did not receive oxygen. Overall, 29,555 patients (90.7%) survived till hospital discharge. In the multivariable unconditional logistic regression, POA had an odds ratio (OR) of 0.33 (95% confidence interval [CI], 0.30–0.37; P <.001) for favorable in-hospital mortality. Furthermore, there were significant differences in all the important variables between the POA and no POA groups (P <.001); therefore, we used propensity score matching analysis. After adjustment for the covariates of selected variables, we found that POA was not associated with a higher rate of survival after hospitalization (adjusted OR, 1.02; 95% CI, 0.99–1.04; P = .27). Even after adjustment for all covariates, POA did not improve in-hospital mortality (adjusted OR, 1.01; 95% CI, 0.99–1.03; P = .08).In this study, POA did not improve in-hospital mortality in trauma patients. However, further studies are needed to validate our results.