Background: We aimed to analyze in-hospital timing and risk factors for mortality in a level 1 trauma center. Methods: This is a retrospective analysis of all trauma-related mortality between 2013 and 2018. Patients were divided and analyzed based on the time of mortality (early (≤48 h) vs late (>48 h)), and within different age groups. Multivariate regression analysis was performed to predict in-hospital mortality. Results: 8624 trauma admissions and 677 trauma-related deaths occurred (47.7% at the scene and 52.3% in-hospital). Among in-hospital mortality, the majority were males, with a mean age of 35.8 ± 17.2 years. Most deaths occurred within 3–7 days (35%), followed by 33% after 1 week, 20% on the first day, and 12% on the second day of admission. Patients with early mortality were more likely to have a lower Glasgow coma scale, a higher shock index, a higher chest and abdominal abbreviated injury score, and frequently required exploratory laparotomy and massive blood transfusion ( P < .005). The injury severity scores and proportions of head injuries were higher in the late mortality group than in the early group. The severity of injuries, blood transfusion, in-hospital complications, and length of intensive care unit stay were comparable among the age groups, whereas mortality was higher in the age group of 19 to 44. The higher proportions of early and late in-hospital deaths were evident in the age group of 24 to 29. In multivariate analysis, the shock index (OR 2.26; 95%CI 1.04-4.925; P = .04) was an independent predictor of early death, whereas head injury was a predictor of late death (OR 4.54; 95%CI 1.92-11.11; P = .001). Conclusion: One-third of trauma-related mortalities occur early after injury. The initial shock index appears to be a reliable hemodynamic indicator for predicting early mortality. Therefore, timely hemostatic resuscitation and appropriate interventions for bleeding control may prevent early mortality.