Abstract:The health benefits of walking and cycling to and from school, also called active school transportation (AST), are well documented. In the context of a declining trend in AST across the Western world, this paper examines school-travel behavior of 11-year-old children in Toronto, using multiple cross-sectional data from 1986, 1996, and 2006 Transportation Tomorrow Surveys. Results from binomial logit models suggest that school-travel distance and neighborhood built environment indeed explain some variation in the odds of AST between 1986 and 2006, and that the correlates of AST may have changed over time. Higher neighborhood block density correlated with walking/cycling in 1986. In contrast, household automobile ownership was negatively associated with AST in 2006; the effect of the built environment was relatively weak for that year. In addition, fewer children walked/cycled in 2006 compared to 1986, even when distance to school was short (<0.8 kilometers). Policy and programs should recognize the potentially changing role of travel distance to school and automobile ownership on a child's school travel outcome. Interventions in neighborhoods with high automobile ownership should specifically focus on education and encouragement to increase AST rates.