Why is it that parents with more schooling tend to have children with better outcomes? We use unique Danish administrative data for identical and fraternal twin parents and their children to estimate the effect of parental schooling on short-run and long-run outcomes for their children. By differencing within identical twin pair we are able to take heritable endowments transmitted from parent to child into account. For all outcomes OLS is found to be upward biased. Father schooling is found to have no causal effect on infant and early childhood health. Mother schooling increases birth weight and the probability of high school completion. For older cohorts, we are able to replicate the findings of Behrman & Rosenzweig (2002) that fathers’ schooling has a positive causal effect on child schooling but mothers’ does not. However, this is reversed for parents born after 1945, when mothers’ schooling has a positive causal schooling effect but fathers’ no longer does.