The hypothesis that alpha-adrenergic coronary vasoconstriction helps maintain a uniform transmural distribution of myocardial blood flow during exercise was tested in dogs. Carotid artery loops were surgically constructed and a splenectomy performed three weeks prior to study. On the day of study, the dog was anesthetized briefly (fentanyl and nitrous oxide) for percutaneous catheterization, and alpha-receptors in one myocardial region were blocked with phenoxybenzamine (0.25 mg/kg) infused selectively into the left circumflex coronary artery. Recirculation of phenoxybenzamine was minimized by drainage of coronary sinus outflow during the infusion. After the dog recovered from the anesthesia, regional blood flow was measured at rest and during graded treadmill exercise with the microsphere technique calibrated by reference blood samples. Average transmural flow was limited by alpha-vasoconstriction and was less in the region where alpha-receptors were intact than in the region where they were blocked, as has been described by others. The ratio of inner layer myocardial blood flow to outer layer flow was better maintained in the region with alpha-receptors intact than in the region with alpha-receptors blocked when myocardial oxygen consumption was 150 microliter/min/g or greater (p less than 0.001). Even though average transmural flow was limited by alpha-receptor activation, inner layer myocardial blood flow was greater in the region with alpha-receptors intact than in the region with alpha-receptors blocked when myocardial oxygen consumption was 500 microliter/min/g or more (p less than 0.05). In conclusion, adrenergic coronary vasoconstriction mediated by alpha-receptors helps to maintain a uniform transmural distribution of myocardial blood flow during exercise in spite of limiting average transmural flow.