Data from a sample of more than 16,000 children born in the United Kingdom in 1958 were studied to identify factors that may affect the risk of having a traffic injury. Five sets of risk factors were examined: physical, developmental, educational, behavioral, and family. Information about these factors were obtained systematically from parents, teachers, and physicians when the children were 7 and 11 years of age. The results were related to traffic injuries occurring for the first time during each subsequent 4-year period. Between 8 and 11 years of age, 431 children had a traffic injury requiring medical attention, and between 12 and 16 years the number was 590. Logistic regression analyses identified a small number of factors associated with injuries, which varied according to the age and gender of the child. When all these factors were entered into a final model, only five remained: fidgety, abnormal behavior, and three measures of family disruption or disadvantage--crowding, family problems, and being removed from the family and placed in the care of the local authority. These findings suggest that it may be unwise to place much reliance on "high risk" preventive strategies by measures of this kind. Instead, more emphasis should be placed on community-based passive and environmental strategies.