The problem with most intergenerational mobility estimates is that unmeasured and inherited abilities prevent us from drawing inferences. In this paper we estimate the intergenerational mobility of schooling and exploit differences between adopted and own birth children to obtain genetically unbiased estimates. Our results provide a much better insight on whether parents (and policy makers) can actually stimulate the educational attainment of future generations. Controlling for inherited abilities and assortative mating we find that the association between mother's (but not father's) and child schooling disappears.