The pharmacokinetic characteristics of a constant rate methohexitone infusion were studied in young ASA 1 patients undergoing maxillofacial surgery. They were randomly assigned to two groups; group M patients (n = 7) were given 9 mg.kg-1.h-1 of methohexitone for one hour, and group MF patients (n = 7) 9 mg.kg-1.h-1 of methohexitone with 7 micrograms.kg-1.h-1 of fentanyl, also for one hour. Blood samples for determining methohexitone concentrations were obtained at various times, from before the start of the methohexitone infusion up to 19 h afterwards. In twelve patients, a two-compartment model was appropriate to characterize the decrease of methohexitone concentration; for the other two (one in each group), a three-compartment model was applied. There were no statistically significant differences between the two groups. Elimination half-life in group M was 3.22 +/- 1.96 h, and total plasma clearance 8.54 +/- 2.8 ml.kg-1.min-1. The wide variations in pharmacokinetic parameters between subjects may explain some unpredictable variations in duration of action of methohexitone. Fentanyl did not modify methohexitone pharmacokinetics, which remained of the first order. However, it potentiated the barbiturate's action: extubation was only possible after stopping the infusion for 39.4 min +/- 22 min in group MF, and 15.4 min +/- 6 min in group M (p less than 0.01). At that time, plasma concentrations were respectively 3.12 +/- 0.99 mg.l-1 (group MF) and 5.71 +/- 2.09 mg.l-1 (group M), (p less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)