Child obesity in the U.S. is a significant public health issue, particularly among children from disadvantaged backgrounds. Thus, the roles of parents' human and financial capital and racial and ethnic background have become important topics of social science and public health research on child obesity. Less often discussed, however, is the role of family structure, which is an important predictor of child well-being and indicator of family socioeconomic status. The goal of this study, therefore, is to investigate how preschool aged children's risk of obesity varies across a diverse set of family structures and whether these differences in obesity are moderated by family poverty status and the mothers' education. Using a large nationally representative sample of children from the Early Childhood Longitudinal Study -- Birth Cohort, we find that preschoolers raised by two biological cohabiting parents or a relative caregiver (generally the grandparent) have greater odds of being obese than children raised by married biological parents. Also, poor children in married biological parent households and non-poor children in married step parent households have greater obesity risks, while poor children in father only, unmarried step, and married step parent families actually have lower odds of obesity than children in non-poor intact households. The implications of these findings for policy and future research linking family structure to children's weight status are discussed.