Increased heart rate and left ventricular pressure during humoral and neuronal adrenergic activation act to restrict blood flow preferentially in the subendocardium. The hypothesis was advanced that alpha-adrenergic coronary vasoconstriction preferentially in the subepicardium may counterbalance the enhanced extravascular compression in the subendocardium and serve to maintain blood flow transmurally uniform. In 40 anesthetized dogs, regional myocardial blood flow was determined with colored microspheres; wall function, with sonomicrometry. Humoral adrenergic activation (HAA) was induced by a combination of intravenous atropine, intravenous norepinephrine, and atrial pacing during baseline coronary vasomotor tone (group 1, n = 6) and in the presence of maximal coronary vasodilation with intravenous dipyridamole (group 2, n = 6). In an additional group, HAA was induced by intravenous norepinephrine in the presence of dipyridamole but without atropine and atrial pacing in order to increase end-diastolic left ventricular pressure (group 3, n = 6). Measurements were performed at rest, during HAA, and during ongoing HAA with the intracoronary infusion of the alpha-antagonist phentolamine (Phen). At unchanged mean aortic pressure, Phen improved blood flow particularly to the inner layers as follows: from 1.42 +/- 0.40 (mean +/- SD) to 1.90 +/- 0.40 mL/(min.g) (group 1, P < .05), from 4.99 +/- 2.31 to 5.53 +/- 2.56 mL/(min.g) (group 2, P < .05), and from 6.01 +/- 1.41 to 6.29 +/- 1.27 mL/(min.g) (group 3, P < .05), associated with a decrease in outer layer blood flow in groups 2 and 3. In 16 additional dogs, beta-adrenoceptors were blocked by propranolol and muscarinic receptors by atropine. Neuronal adrenergic activation (NAA) was induced by cardiac sympathetic nerve stimulation (CSNS) during baseline coronary vasomotor tone (group 4, n = 8) and in the presence of maximal vasodilation (group 5, n = 8). Measurements were performed at rest, during a first CSNS, and 20 minutes later during a second CSNS+Phen. The reproducibility of two consecutive episodes of CSNS 20 minutes apart was demonstrated in a separate set of experiments (n = 6). At matched mean aortic pressures, Phen improved blood flow to all myocardial layers in group 4, whereas in group 5, Phen induced a redistribution of myocardial blood flow toward subepicardial layers [from 4.44 +/- 0.96 to 4.81 +/- 0.83 mL/(min.g), P < .05] at the expense of inner layers. With the addition of Phen, there was no change in regional wall function in any group of dogs studied. Thus, during HAA, alpha-adrenergic coronary vasoconstriction does not exert a beneficial effect on transmural blood flow distribution. During NAA, a beneficial effect of alpha-adrenergic coronary vasoconstriction becomes apparent only under conditions of maximal coronary vasodilation.