A national register-based cohort study was conducted to explore whether the social patterning of car drivers suffering injury repetitions differs from that of once-injured drivers in a cohort of young persons in Sweden (aged 18-26). Injury repeaters were defined as individuals sustaining injuries as a car driver on more than one occasion over an 8-year period. Only subjects obtaining a driver's licence before the age of 27 were included in the study. The social variables considered were, in turn, gender, socioeconomic position of origin, and own educational attainment. Attention was also paid to age at licensing. Two types of comparisons were made, using odds ratios with 95% confidence intervals. First, the odds of injury repeaters were computed using the once-injured car drivers as a reference group. Second, odds were compiled for once, twice, and three or more times injured people, using the non-injured at all as a comparison group. The results show that, by and large, the injury-risk distribution of car drivers with injury repetition does not differ from that of one-injury drivers with regard to gender, education, or socioeconomic group. However, drivers from self-employed households show greater odds of injury repetition (OR 1.65, CI 1.02-2.67) than of one injury compared with drivers from the families of non-manual employees. Since the number of injury repeaters is low and their socioeconomic distribution is very similar to that of the once injured, there is no need to regard them as a group at excess risk or different from the one-time injured. Reducing risk levels and risk differentials for the one time injured should therefore receive priority with regard to traffic-injury prevention.