The authors examined direct myocardial and coronary vascular responses to the anesthetic induction agents etomidate, ketamine, midazolam, propofol, and thiopental and compared their effects on attenuating autoregulation of coronary flow as assessed by changes in oxygen supply/demand relationships. Spontaneous heart rate, atrioventricular conduction time during atrial pacing, left ventricular pressure (LVP), coronary flow (CF), percent oxygen extraction, oxygen delivery, and myocardial oxygen consumption (MVo2) were examined in 55 isolated guinea pig hearts divided into five groups of 11 each. Hearts were perfused at constant pressure with one of the drugs administered at steady-state concentrations increasing from 0.5 microM to 1 mM. Adenosine was given to test maximal CF. At concentrations below 10 microM no significant changes were observed; beyond 50 microM for midazolam, etomidate, and propofol, and 100 microM for thiopental and ketamine, each agent caused progressive but differential decreases in heart rate, atrioventricular conduction time (leading to atrioventricular dissociation), LVP, +dLVP/dtmax, percent oxygen extraction, and MVo2. The concentrations (microM) at which +dLVP/dtmax was reduced by 50% were as follows: etomidate, 82 +/- 2 (mean +/- SEM); propofol, 91 +/- 4; midazolam, 105 +/- 8; thiopental, 156 +/- 11; and ketamine, 323 +/- 7; the rank order of potency was etomidate = propofol = midazolam greater than thiopental greater than ketamine; results were similar for LVP. At the 100 microM concentration, CF was decreased 11% +/- 2% by ketamine and 5% +/- 3% by thiopental but was increased 17% +/- 6% by etomidate, 21% +/- 5% by midazolam, and near maximally to 57% +/- 10% by propofol; MVo2 was decreased 8% +/- 4% by thiopental, 10% +/- 5% by ketamine, 19% +/- 5% by midazolam, 29% +/- 7% by etomidate, and 37% +/- 5% by propofol; oxygen delivery/MVo2 was unchanged by thiopental and ketamine but was increased 62% +/- 7% by midazolam, 71% +/- 9% by etomidate, and 150% +/- 15% by propofol. Between 100 microM and 1 mM, thiopental and ketamine did not increase CF but decreased MVo2 and percent oxygen extraction, whereas propofol maximally increased CF and decreased MVo2 and midazolam and etomidate had intermediate effects. These results indicate that on a molar basis, propofol, and less so midazolam and etomidate, depress cardiac function moderately more than thiopental and ketamine, and that propofol markedly attenuates autoregulation by causing coronary vasodilation. With doses used to induce anesthesia, propofol and thiopental appear to depress cardiac function more than ketamine or etomidate.