A substantial and growing fraction of children across Europe and the US live in single parent households. Law practices are evolving to encourage both parents to maintain contact with their children following parental separation/divorce, driven by the belief that such contact is in the best interest of the child. We test this assumption by using information on the distance between non-residential parents and their children to proxy for contact, and measuring educational, behavioral, and health outcomes for a population sample of children from nonnuclear families in Denmark. Instrumental variables techniques are employed to control for the endogeneity of residence. The results indicate that educational and behavioral outcomes are better for children who live farther away from their non-residential parent, but that distance is not related to health outcomes. Failing to control for endogeneity biases the results in favor of more proximate parents. These findings suggest that policy efforts to keep separated parents geographically closer together for the sake of the children may, in fact, not be advantageous.