This study was undertaken to assess the effect of collateral circulation on myocardial blood flow (MBF) reserve (ability to increase myocardial blood flow with a stress that increases myocardial oxygen requirements). One hundred patients had MBF measured at rest and after isoproterenol. After classification by anatomic severity of coronary artery disease (CAI), the groups were further compared by presence or absence of collateral vessels, and whether the collaterals were intercoronary or bridge collaterals. Forty patients (group A) had no coronary artery disease demonstrated by cineangiography. The increase in MBF with isoproterenol for this group was 87% (P< 0.001). Fifteen patients (group B) had CAI of 175 or greater. These patients increased MBF 73% on infusion of isoproterenol, an insignificant difference from group A. Forty-five patients (group C) had CAI of 175 or less when an index of 300 represents no occlusive disease. Those with intercoronary collateral vessels (group C1, 2) were unable to increase MBF to the same extent as patients in groups A and B did. There was no difference between this group and those without collateral vessels and the same severity of disease. Fifteen patients with CAI of 175 or less (group C3, 4) had bridge collaterals and were able to increase MBF to a greater extent than those with no collaterals or with intercoronary collateral vessels. This same group of patients, in a parallel observation, showed less S-T-segment depression on treadmill exercise than patients with intercoronary collateral vessels.The data suggest that intercoronary collateral vessels contribute insignificantly, statistically, to myocardial blood flow reserve. Bridge collaterals, however, do seem to contribute in selected patients.