Because certain vasoactive drugs also possess negative myocardial inotropic properties, it is difficult to determine their direct pharmacological actions on coronary resistance vessels in situ. Thus, a drug's myocardial effects may induce indirect physiological changes in coronary vascular tone that confound measuring its direct action. To separate the direct from indirect drug actions on vascular tone, the authors describe an experimental system utilizing the isolated perfused rat heart arrested with tetrodotoxin. The direct effects of the volatile anesthetics halothane and isoflurane on coronary vascular tone were examined in this preparation, using the concept of minimum alveolar concentration as the measure of potency. Both anesthetics demonstrated a dose-dependent direct coronary vasodilation that was reversible, with the median effective dose being 1.31 +/- 0.08 (mean +/- S.E.) for halothane, and 1.53 +/- 0.12 for isoflurane (P = 0.06; units of measure are the fraction of minimum alveolar concentration for each anesthetic). Further studies indicated that maximal vasodilation by adenosine was diminished after 90 min of perfusion, independent of anesthetic administration. Myocardial oxygen consumption was decreased significantly from the beating state by arrest, but neither anesthetic nor adenosine affected myocardial oxygen consumption further. These data indicate that halothane and isoflurane are equipotent for inducing direct dilation of coronary resistance vessels in the isolated perfused arrested rat heart.