Using the 1979 National Longitudinal Survey of Youth along with the child survey, we examine the relationship between urban sprawl of U.S. metro counties and the body mass index (BMI) of children who reside in these counties. We make a distinction between urban sprawl in a county and its geographical placement in the urban hierarchy. Even after accounting for unobserved individual heterogeneity and resulting selection bias, we find that urban sprawl is positively related to child BMI and distance to large metros is negatively related to child BMI. These effects are somewhat pronounced among girls and middle/high school children.