Data on host factors influencing mortality in trauma patients is sparse and contradictory. To develop a model for health policy decisions, we examined all trauma admissions to acute care hospitals in the state of California in the year 1986. We looked at the influence of the following host factors: age, gender, and preinjury medical conditions, on mortality stratified by injury severity. The study group (N = 199,737) had an overall mortality rate of 1.9%. Mortality increased starting at age 40 years and was independently influenced by gender, the presence of pre-existing disease, and the body region injured. In patients with minor injury, mortality rates became higher in the elderly at age 65+. However, in patients with injuries of moderate severity, mortality increased in both middle age (40-64) and elderly groups (65+). Male gender was also a risk factor, present in both the elderly and middle age groups. While the presence of both pre-existing medical disease or injury to head or abdomen was related to increased mortality in middle-aged patients at all severity levels, neither accounted for the effect of gender. Conclusion. Age and gender influence mortality in trauma patients. These effects are not explained by documented pre-existing disease or region of injury. Age and gender serve only as observable markers for subgroups of patients with impaired response to injury. Middle-aged males comprise a previously unrecognized high-risk subgroup for this impaired response.