This study tested whether ATP-dependent K+ channels (KATP channels) are an important mechanism of functional coronary hyperemia in conscious, instrument-implanted diabetic dogs. Data were collected at rest and during exercise before and after induction of diabetes with alloxan monohydrate (40-60 mg/kg intravenously). KATP channels were inhibited with glibenclamide (1 mg/kg intravenously). In nondiabetic dogs, arterial plasma glucose concentration increased from 4.8 ± 0.3 to 21.5 ± 2.2 mmol/l 1 week after alloxan injection. In nondiabetic dogs, exercise increased myocardial oxygen consumption (MVo2) 3.4-fold, myocardial O2 delivery 3.0-fold, and heart rate 2.4-fold. Coronary venous Po2 decreased from 19.9 ± 0.8 mmHg at rest to 14.8 ± 0.8 mmHg during exercise. Diabetes significantly reduced myocardial 02 delivery and lowered coronary venous Po2 from 16.3 ± 0.6 mmHg at rest to 13.1 ± 0.9 mmHg during exercise. Glibenclamide did not alter the slope of the coronary venous PO2 versus MVo2 relationship in nondiabetic dogs. In diabetic dogs, however, glibenclamide further reduced myocardial O2 delivery; coronary venous Po2 fell to 9.0 ± 1.0 mmHg during exercise, and the slope of the coronary venous PO2 versus MVo2 relationship steepened. These findings indicate that KATP channels contribute to local metabolic coronary vasodilation in alloxan-induced diabetic dogs.