In this paper I want to examine the theoretical foundations of "monetarism". At the outset there is a difficulty. Professor Stein writes: "Monetarists are policy oriented. Their major propositions are a series of empirical observations. . . rather than a theory in direct opposition to neo-Keynesian analysis" (Stein, 1976). And Professor Friedman writes, "I continue to believe that the fundamental differences between us are empirical not theoretical" (Friedman, 1976). Professor Friedman also urges us to paint with a broad Marshallian brush and not to worry too much about the fine detail. All this suggests the claim that there is an accepted theory of the economy and that this theory is capable of yielding both monetarist and other conclusions. For instance, in the IS-LM context the LM curve may be vertical and/or the IS curve could be flat, in which case Monetarists' propositions hold. But the shapes of these curves are cdntingent on time and place, and Monetarists do not claim that logically the curves could not have less extreme shapes. Yet it will readily be agreed that this classification of cases by the shapes of the two curves does not exhaust the propositions of either Monetarists or nonMonetarists. For instance, the former will want to argue that in normal times the intersection of the two curves is at a level of income producible by the economy with a natural rate of unemployment. Moreover, it is by no means the case that we are agreed that the IS-LM cross is a generally accepted theory of the economy. So for the moment at least let us suspend judgment on this matter. There is also another difficulty: how are we to define monetarism? Professor Friedman himself has drawn our attention to the necessarily coarse division of opinions any definition is bound to imply (Friedman, 1976). I think he is right and accordingly do not attempt a formal definition. It should be realized that nothing of substance hinges on a definition. I shall, as the story unfolds, label certain views as monetarist, and I shall naturally not be wilfully perverse in this. But it will be of no importance whether there exists an individual who assents to all the propositions thus labelled. I am not here interested in the history of thought.