I begin with the view that the usual property‐based epiphenomenalist challenges to anomalous monism are unconvincing in light of Davidson's reluctance to analyze causation in terms of properties. I argue, however, that the challenges against Davidson do hold in the weaker sense that although mental events have causal efficacy the identification of an agent's reasons does not causally explain behaviour. I then show that in light of Davidson's commitment to psychophysical supervenience this does not constitute a serious problem for anomalous (...) monism. (shrink)