To determine the site of isoflurane-associated coronary vasodilatation, the authors measured epicardial coronary artery diameter and examined the effects of isoflurane on coronary arteriolar tone. Angiograms of the left coronary system were obtained in seven fentanyl-pentobarbital anesthetized dogs and quantitated with a computerized analysis system. Cross-sectional areas of the proximal, mid, and distal left anterior descending and proximal circumflex coronary arteries were obtained at three arterial pressures, and then the measurement repeated following administration of 0.75%, 1.5%, and 2.25% end-tidal isoflurane. At the same time coronary blood flow was measured using a 123Xe washout technique. Isoflurane was found to have no effect on epicardial coronary artery dimensions. No dilatation was observed throughout the range of isoflurane concentrations and coronary perfusion pressures investigated. However, despite the absence of epicardial coronary effects, coronary arterioles were dilated by both 1.5% and 2.25% isoflurane. Coronary blood flow corresponding to a myocardial oxygen consumption of 7.5 ml oxygen.100 gm−1.min−1 was calculated as 99 ± 17 ml.100 gm−1.min−1 (mean ± SD) during control conditions, and it increased to 150 ± 26 ml.100 gm−1.min−1 at 1.5% isoflurane (P < .004) and to 197 ± 35 ml.100 gm−1.min−1 at 2.25% isoflurane (P < .001). Although higher concentrations of isoflurane dilated intramyocardial arterioles, isoflurane had no effect on epicardial coronary arteries.