SUMMARY The resistance to coronary blood flow in various parts of the myocardium was studied with the tracer mkrospberes technique before and immediately after an acute coronary occlusion and several weeks after a more slowly occurring coronary occlusion by Ameroid constrictor. All experiments were carried out in the isolated, metabolically supported, empty, beating dog heart at maximal coronary vasodilation induced with adenosine. Coronary resistance of the normal empty beating heart at maximal coronary vasodilation was 0.20 mm mm Hg/(ml/min) per 100 g of tissue (subepicardium) and 0.16 mm Hg/(ml/min) per 100 g of tissue (subendocardium). After acute coronary occlusion the perfusion of the subtended myocardium was maintained at a much lower level by way of collateral vessels, which snowed i resistance to flow of 3.52 mm Hg/(ml/miD) per 100 g. If coronary artery occlusion proceeded more slowly the collateral vessels became more functional and myocardial infarction was avoided. During collateral enlargement collateral resistance fell from 3.52 to 0.22 mm Hg/(ml/min) per 100 g within a period of 8 weeks after implantation of the constricting device. The degree of compensation by collaterals for the loss of the occluded native coronary artery was 33% of its former conductance.CORONARY ARTERY occlusion without myocardial infarction is a common finding at autopsy in humans; in reported cases'* collateral vessels had enlarged quickly enough and to a sufficient degree to prevent irreversible myocardial damage. Coronary arterial lesions, however, are not always compensated for by collateral enlargement, and clinical and technical limitations have made it impossible to study the collateral circulation quantitatively. Methodological limitations are also responsible for the fact that the collateral resistance in acute and chronic experimental coronary occlusion remained unknown.Until recently the time course of collateral development was estimated by the use of indirect indices because collateral flow proper, hence collateral resistance, could not be measured. The introduction of the tracer microsphere method into circulation research, 5 however, has changed this situation and we have tried to study the increase of collateral flow with time, using this method. Since the relationship between perfusion pressure and coronary collateral flow might not be linear and since zero flow must be assumed at positive pressures, i.e., the critical closing pressure,* collateral flows and pressures were measured over a wide physiological range. In this way the resistance of the collateral vessels as well as the degree of compensation for the loss of the native circulation, was quantitatively determined.