Although traffic accidents constitute a major health problem among North American children, the extent of this problem has largely been defined by means of police reports. In order to more accurately determine the incidence of motor vehicle accidents (MVA's) among children, an epidemiologic survey was conducted over an 18-month period in one metropolitan area (population 1.76 million). All passenger, pedestrian and cyclist injuries occuring to children age 0-14 years requiring medical attention were identified through a monitoring system established in the emergency rooms of 11 hospitals scattered throughout the city. Information regarding age, sex, area of residence and extent of injury was recorded for each case. In addition, police reports which provided details of the accident itself, and death certificates were obtained and linked by computer with existing hospital data. The police reports also served to ensure that no victims had been overlooked. Any cases which failed to link were reviewed and, when possible, medical records were retrieved from the appropriate hospital.

Over the 18-month period a total of 1,767 children were injured in traffic accidents. Of these 27% were passengers, 60% were pedestrians and 14% were cyclists. By type of accident, 13%, 33% and 28% of passengers, pedestrians and cyclists respectively suffered moderate to severe injuries i.e. MAIS≥2. Incidence rates by age and sex reveal that boys age 5-9 year and 10.-14 years experience almost twice as many pedestrian accidents as girls;​ this ratio is the same for both minor and serious injuries. Demographic information shows geographic clusterings of pedestrian and cyclist accidents suggesting an association with low SES municipalities. A comparison of hospital and police reports for 1981 shows that 39% of the children involved in MVA's were not reported to the police. Of these almost 30% had an MAIS≥2.