The hemodynamic, electroencephalographic (EEG), and metabolic effects of a high-dose methohexital anesthetic were examined in eight neurosurgical patients. The patients were studied at rest and at 15-min intervals during a 60-min infusion of the drug, given at a rate of 0.40 mg·kg-1·min-1 (total dose 24 mg/kg). Ventilation was controlled with oxygen:air (FI(O2) = 0.50), and fluid was infused at a rate sufficient to maintain pulmonary capillary wedge (PCW) pressures at control values (8 ± 2 mmHg, mean ± SD). Serum methohexital concentrations increased progressively, reaching values of 11.7 ± 2.9 μg/ml at t = 30 min and 18.1 ± 10.8 μg/ml at t = 60 min. Characteristic barbiturate-induced EEG changes were noted, with isoelectricity achieved at t = 28 ± 13 min. Methohexital infusion resulted in significant reductions in arterial pressure (84% of control at t = 60 min), systemic vascular resistances (83% of control at t = 60 min), right and left ventricular stroke work indices (65% and 68% of control, respectively at t = 60 min), and total body O2 consumption (76% of control at t = 60 min). In addition, a progressive dose-related decrease in stroke volume index was noted (50.1 ± 90 ml.beat-1·m-2 at t = 0, 40.1 ± 10.2 ml·beat-1·m-2 at t = 60 [80% of control]). This occurred in spite of unchanged ventricular filling pressures. However, cardiac index was well maintained (unchanged at t = 60 min) because of increases in heart rate (123% of control at t = 60 min). There was no change in Pa(O2), Pa(CO2), or pulmonary vascular resistance. These data demonstrate that doses of methohexital sufficient to produce profound EEG suppression are accompanied by both vasodilation and some depression of myocardial function, even when ventricular filling pressures are maintained. Nevertheless, the magnitude of these changes suggests that high doses of methohexital may be a hemodynamically acceptable form of anesthesia for certain restricted neurosurgical procedures. However, refractory postoperative seizures occurring in three patients indicate that this anesthetic technique has potentially serious associated difficulties. For this latter reason, the authors have suspended their use of methohexital and are examining the utility of alternative barbiturates.