The National Accident Sampling System (NASS) was analyzed to provide quantitative information on injuries to collision-involved automobile occupants in relation to age. The analysis shows differences between age groups in the following characteristics of an occupant's injuries: the number of injuries, pattern of injury to different body parts, kinds of injuries sustained, and severity of injury. The analysis also shows that the description of outcomes such as use of outpatient medical services, hospitalization, and occurrence of death would be better accomplished from hospital-based data. The NASS data are at best of limited usefulness in trying to establish this kind of comparison. Limitations of the NASS data base for descriptive epidemiologic research are defined, and recommendations for research directions are presented.