The costs of inpatient hospital care for treating head trauma were examined in 125 patients over the age of 50. Regression analysis indicated that injury severity, as measured by the admitting Glasgow Coma Scale, was a good predictor of hospital cost, while patient age was unrelated to cost. The regression relationship, however, was quadratic (not linear), indicating that least injured and severely injured patients cost less than moderately injured patients. These findings are in direct contrast with the commonly held belief that the elderly consume more hospital resources than younger patients. They suggest the need for a reexamination of the use of age in setting prospective reimbursement amounts in certain diagnoses, and in making resource allocation decisions relative to geriatric programming at the hospital level. Head trauma patients are costly because they are sick, not because they are old.