The purpose of this study was to examine the effects of cardiac sympathetic nerve stimulation on regional coronary blood flow following .beta.-blockade. In 17 anesthetized dogs treated with propranolol (2mg/kg iv) regional myocardial perfusion was measured (microspheres) during control and during stimulation of the ventrolateral, ventromedial, or recurrent cardiac nerve (8-10 V, 4-ms pulses at 10 Hz for 30 s). Ventrolateral nerve stimulation produced 25.5 .+-. 3.4 and 23.5 .+-. 3.1% (mean .+-. SE) decreases in coronary blood flow in the posterior and lateral quadrants of the left ventricle. These changes were significantly greater than the 8.5 .+-. 2.9, 11.5 .+-. 3.0, and 3.7 .+-. 2.8% decreases in the anterior and septal left ventricle and right ventricle, respectively (P < 0.01). Ventromedial nerve stimulation produced 18.6 .+-. 2.8, 15.4 .+-. 2.8, and 10.1 .+-. 3.2% decreases in flow in the anterior, septal, and lateral left ventricle, respectively. These changes were significantly greater than the 5.3 .+-. 3.8 and 9.9 .+-. 3.6% decrease in the posterior left ventricle and right ventricle (P < 0.01). Recurrent cardiac nerve stimulation produced 16.4 .+-. 2.1, 15.6 .+-. 2.2, and 13.6 .+-. 2.5% decrease in flow in the anterior and septal left ventricle and right ventricle, respectively. These changes were significantly greater than the 5.2 .+-. 3.2 and 5.4 .+-. 3.0% changes in the posterior and lateral quadrants (P < 0.01). Ventrolateral nerve stimulation resulted in a small but significant increase in the endocardial-to-epicardial blood flow ratio in the posterolateral left ventricle. We conclude that the cardiac sympathetic nerves have selective regional effects myocardial perfusion that could contribute to fine regulation of flow in the normal heart and could have beneficial or adverse effects in the setting of coronary stenosis or occlusion.