Tobin and Brainard have developed a series of models of the financial sector in which income, expectations and the stocks of certain assets are exogenously given, and in which asset markets are assumed to be in equilibrium (Tobin and Brainard, 1963; Brainard, 1967; Tobin, 1969). These models can be used to analyse the effects of changes in the exogenous variables on asset prices and interest rates. Their results may be regarded as exploring the vertical displacements of the LM curve which result from monetary actions (Brainard, 1967, p. 99). By assuming that financial markets adjust virtually instantaneously, such results describe the very short-run, or impact, effects of policy actions: the analysis is of a temporary equilibrium established after a policy change, and before output or expectations respond. This very short run is exactly the situation where credit rationing would be expected, for while the prices of financial assets can be assumed to respond immediately to any discrepancy between demand and supply, the same cannot be said of the lending and borrowing rates offered by financial institutions. If, for example, the rate of interest on bank loans does not rise immediately in response to a contractionary monetary policy, there will be excess demand for loans and credit rationing. When it comes to analysing credit rationing, partial equilibrium analysis using supply and demand curves for loans can be used to draw two main conclusions (see Chick, 1973, pp. 62-67): that raising the loan rate will increase the supply of loans and thus be expansionary; and that a restrictive monetary policy will have a greater impact if loan rates are sticky, shifts in the supply curve not being offset by movements along it. Two problems arise with such partial equilibrium analysis: the supply curve may be inappropriate given that the supply of loans depends not only on bank behaviour but also on the public's preferences; and the relationship between the volume of lending, the loan rate and aggregate demand is not clear, for with rationing lending is restricted but the cost of loans for those who have them is lower than in the absence of rationing. Both problems can be circumvented by considering credit rationing in the context of a Tobin-Brainard type of general equilibrium model. This paper is concerned with formulating such a model and using it to answer the following questions: (a) if there is rationing, will raising the loan rate be expansionary or contractionary? (b) will open-market operations be more or less effective if the loan rate is sticky? and (c) what effect will a sticky loan rate have on the slope of the LM curve? Of these questions the first is fundamental, for the other questions can be answered by dividing the effects of a policy change into (a) the effects of a policy change given a fixed loan rate and (b) the effects of changing the loan rate to its new equilibrium level. So the bulk of the paper is given over to discussing the first question. Subscripts on all letters other than r denote partial differentiation.