Direct vasodilation of coronary resistance vessels by anesthetics may reduce coronary flow reserve and interfere with myocardial flow-metabolism coupling. This study was performed to evaluate the potential for the halogenated anesthetic agents sevoflurane, isoflurane, and halothane to alter the regulation of coronary flow via a direct action on coronary resistance vessels. Coronary flow and flow reserve were measured in the quiescent isolated perfused rat heart at anesthetic concentrations between 0 and 3 x MAC. In order to minimize anesthetic-induced secondary changes in coronary resistance, constant coronary perfusion pressure was maintained; the left ventricular cavity was vented; and tetrodotoxin was used to achieve cardiac arrest. These conditions permitted the dissociation of direct anesthetic actions from indirect regulatory processes affecting coronary vascular resistance (CVR). Coronary flow reserve was defined as the difference between coronary flow prior to and during administration of a maximally vasodilating dose of adenosine. Each anesthetic significantly reduced the magnitude of both CVR and coronary flow reserve in a concentration-dependent manner. Sevoflurane reduced coronary flow reserve significantly less than did halothane and isoflurane. At high concentrations (3.0 x MAC), coronary flow reserve was abolished by halothane and was decreased to near zero by isoflurane; however, flow reserve was reduced only 48% from control by sevoflurane. This difference among anesthetics is explained primarily by variations in the magnitude of direct coronary vasodilation produced by each anesthetic, rather than by effects on maximal vasodilator capacity. These data show that sevoflurane's intrinsic vasodilator action on coronary resistance vessels differs substantially from that of halothane and isoflurane.