This paper is about a puzzle which lies at the heart of contemporary physicalist theories of mind. On the one hand, the original motivation for physicalism was the need to explain the place of mental causation in the physical world. On the other hand, physicalists have recently come to see the explanation of mental causation as one of their major problems. But how can this be? I-low can it be that physicalist theories still have a problem explaining something which their physicalism was intended to explain in the first place? lf physicalism is meant to be an explanation of mental causation, then why should it still face the problem of mental causation? Disentagling this puzzle will cast light both on the recent mental causation debate and on physicalism itself. We can make a broad distinction between those forms of physicalism which identify mental and physical items and those which claim that there is some weaker relation of ’constitution’ between the mental and the physical. This latter view is now the orthodox version of physicalism. I shall argue that the problem of mental causation is only a problem for this orthodox physicalism, and not for identity theories. ln itself, this is not a particularly unusual claim. But I shall also argue that the real lesson of the mental causation debate is that orthodox physicalism is either unstable or unmotivated. It is unstable because (unlike the identity theories) it cannot reconcile mental causation with its other physicalist assumptions. It is unmotivated because in attempting to solve this mental causation problem, orthodox physicalism typically abandons one (or more) of the assumptions which form part of the original motivation for physicalism. To establish this, I need to explain (a) the nature of the arguments for physicalism, (b) the problem of mental causation, and (c) the standard solutions to the problem. These three tasks will form the main substance of this paper. But first I need to make some preliminary remarks about physicalism...