Objectives: To determine the relative risk of mortality after traumatic brain injury (TBI) in elderly and nonelderly populations. Design: Population-based, longitudinal, retrospective study. Setting: Olmsted County, Minn. Participants: From a list of all residents with any diagnosis indicative of potential TBI from 1985 through 1999, we identified a random sample of 7175 cases and reviewed their complete provider-linked medical records to confirm and establish sequence of TBI events. Cases were grouped into 3 age groups, less than 16 years, 16-65 years, and 65 years and more, with 16-65 years used as a reference group. Measurements: Individuals with confirmed incident TBI were categorized as moderate/severe or mild and followed for vital status through June 30, 2004. Short-term (<6 months) and long-term (6 months - 10 years) mortality after TBI was assessed. Observed mortality after TBI was compared with expected mortality, on the basis of rates for the Minnesota white population of similar age and sex, controlling for calendar year of TBI. Results: Of the 1433 confirmed incident TBI cases, 508 were younger than 16 years, 790 were aged 16-65 years, and 135 were aged 65 years and more. In the elderly population, 6-month mortality was 20% compared with 3.4% for 16- to 65-year-olds and 1.2% in those aged less than 16 years. Among 6-month survivors, 10-year mortality was 45.4% in the elderly group compared with 2.6% for 16- to 65-year-olds and 0.2% in those aged less than 16 years. Elderly patients were more likely to die within the first 6 months after TBI with less severe injuries than were nonelderly patients. However, being elderly did not impose additional risk of dying after TBI after controlling for age, sex, and calendar year. Injury severity, not age, was related to increased risk of death. Conclusion: Long-term mortality of TBI patients did not differ between elderly and nonelderly patients when compared to their peers of similar age and sex distribution.