To test if observed vs. expected mortality differs by age among traumatic brain injury (TBI) cases, a population-based, historical cohort study was conducted in Olmsted County, Minnesota. From all residents with any diagnosis suggestive of TBI 1985-1999, we randomly sampled 7,800 and reviewed their medical records to confirm the event. Confirmed incident cases were categorized by age in years (<16 = pediatric, 16-65 = adult, > 65 elderly) and severity (moderate/severe vs. mild) and followed for vital status through 6/30/2004. We compared observed 6-month and 10-year mortality with expected and tested if the differences varied by age. Of 1,433 confirmed incident cases, 35% were pediatric; 55% were adult; only 9% were elderly; 11.2% of all cases were moderate/severe; the proportions by increasing age group were 11.4%, 8.5%, 26.7%. The proportions who died within 6 months increased with increasing age group, both for moderate/severe (10.3%, 40.3%, 50.0%) and mild cases (0%, 0%, 9.1%); mortality for moderate/severe cases was nearly 40 times that for mild cases, independent of age. Among 6-month survivors, 10-year mortality differed from expected only for adult cases. For all cases, after adjusting for sex, year of TBI, and severity, the difference between observed and expected 10-year mortality was greater for adult cases than for pediatric cases and similar for adult and elderly cases. Elderly individuals account for <10% of TBI cases and >50% of 10-year mortality, yet much of this discrepancy reflects age-associated mortality in general. Findings have implications for (1) reducing the number of excess deaths following TBI and (2) caring for survivors.