Treatment of angina pectoris falls within two spheres: (1) increasing myocardial blood flow (direct bypass, vasodilators), or (2) reducing oxygen demand (beta blockade, propranolol). If the latter depresses the stimulus for collateralization without improving input, its use in certain forms of coronary disease may require reconsideration. To examine this hypothesis, 25 dogs (four groups) were evaluated by surface mapping, angiographic collateral mapping, and hemodynamic studies including cardiac output, systemic and coronary pressures, coronary flow, and left ventricular end-diastolic pressure, pressure to time (LVdp/dt). Collateralization was induced with ameroid constrictor placement in Groups I, III, and IV. Groups II and III received immediate propranolol treatment (3 to 5 mg. per kilogram per day for 6 weeks); Group IV received delayed treatment (4 weeks after constriction); and Group I received no propranolol (constrictor alone). Collateral formation was depressed significantly in the acutely beta-blockaded dogs (Group III) by surface and angiographic mapping and by hemodynamic evidence of depressed coronary flow and altered coronary retrograde pressures. Alterations in established collaterals (Group IV) were not consistent. These data suggest that pharmacological reduction of myocardial oxygen demand reduces coronary collateralization experimentally, requiring further evaluation of its usage in multivessel coronary disease to the exclusion of direct bypass.