We report the results of a case-control study designed to identify modifiable risk factors for childhood sledding injuries. Cases included all children injured while sledding (n = 17) during a five-day winter storm who presented for care to a Seattle pediatric emergency department. Two controls were chosen randomly for each case from children who presented to the same emergency department during the ensuing two months for any illness that did not require hospitalization. To qualify as controls, children had to have sledded during the same five-day period but not sustained an injury requiring medical attention. Data were collected on cases and controls using a telephone questionnaire. Analysis of risk factors revealed a significant increase in risk of injury for children who sledded in streets versus those who sledded in a park (OR = 5.1; 95% CI = 1.1-24.1), those who sledded without adult supervision (OR = 5.6; 95% CI = 1.1-26.9), and those whose parents had an annual income > or = $50,000 (OR = 5.7; 95% CI = 1.5-20.8). No difference in risk of injury was found regarding the type of sled used, the number of children, or their position on the sled or for those children with a history of prior sledding experience. We conclude that safe alternatives to sledding in the streets and an increase in adult supervision may reduce the number of childhood sledding injuries.