This study tested whether -adrenoceptor mediated coronary vasoconstriction is augmented during exercise in diabetes mellitus. Experiments were conducted in dogs instrumented with catheters in the aorta and coronary sinus and with a flow transducer around the circumflex coronary artery. Diabetes was induced with alloxan monohydrate (n = 8, 40 mg/kg, iv). Arterial plasma glucose concentration increased from 4.7 ± 0.2 mM in non-diabetic, control dogs (n = 8) to 21.4 ± 1.9 mM one week after alloxan injection. Coronary blood flow, myocardial oxygen consumption (MVO₂), aortic pressure, and heart rate were measured at rest and during graded treadmill exercise before and after infusion of the -adrenoceptor antagonist phentolamine (1 mg/kg iv). In untreated diabetic dogs, exercise increased MVO2 2.7-fold, coronary blood flow 2.2-fold, and heart rate 2.3-fold. Coronary venous PO2 fell as MVO₂ increased during exercise. Following -adrenoceptor blockade, exercise increased MVO₂ 2 3.1-fold, coronary blood flow 2.7-fold, and heart rate 2.1-fold. Relative to untreated diabetic dogs, -adrenoceptor blockade significantly decreased the slope of the relationship between coronary venous PO₂ and MVO₂. The difference between the untreated and phentolamine treated slopes was greater in the diabetic dogs than in the nondiabetic dogs. In addition, the decrease in coronary blood flow to intracoronary norepinephrine infusion was significantly augmented in anesthetized, open-chest,-adrenoceptor blocked diabetic dogs compared to the nondiabetic dogs. These findings demonstrate that-adrenoceptor mediated coronary vasoconstriction is augmented in alloxan-induced diabetic dogs during physiological increases in MVO₂.