We review recent advances in our understanding of placental evolution with particular focus on the interhaemal barrier. It seems likely that the non-invasive, epitheliochorial placentation of living eutherians is a derived state. However, there is disagreement on whether the last common ancestor of eutherian mammals had an endotheliochorial placenta or a haemochorial one. Research has been stimulated by improved understanding of the relations between the orders of mammals provided by molecular phylogenetics. In part, the uncertainties arise from doubt about how to root the mammalian tree. Resolution of this issue will require improved taxon sampling in molecular analyses. At the same time, we need to foster research in comparative placentation on relevant taxa, particularly at the ultrastructural level. Only then can we ensure that information available about the placenta is adequate to capitalise on future advances in molecular phylogenetics. Examples are given of recent findings that could benefit cladistic analysis of placental evolution.