The authors studied the effects of enflurane, isoflurane, and fentanyl, each in combination with 60% nitrous oxide, on the vecuronium infusion rate necessary to maintain constant 90% depression of control muscle twitch tension. Thirty healthy surgical patients were given an initial 0.1 mg/kg bolus of vecuronium, followed by an infusion of vecuronium at an initial rate of 1.0 microgram . kg-1 . min-1. After 1 h of steady-state 90% twitch depression, plasma vecuronium concentrations (Css90) were measured by capillary column gas chromatography. Total plasma clearance of vecuronium was estimated using Css90 values. Vecuronium infusion rates (mean +/- SD) were similar for patients given enflurane (0.28 +/- 0.13 microgram . kg-1 . min-1) and isoflurane (0.30 +/- 0.13 microgram . kg-1 . min-1), but significantly higher in patients given fentanyl (0.92 +/- 0.37 microgram . kg-1 . min-1). Values for Css90 in the patients receiving enflurane and isoflurane were similar (71 +/- 34 and 72 +/- 44 ng/ml, respectively), but significantly higher in those receiving fentanyl (165 +/- 48 ng/ml). Total plasma clearance was similar during enflurane, isoflurane, and fentanyl anesthesia (4.4 +/- 2.6, 4.6 +/- 1.2, and 5.6 +/- 1.9 ml X kg-1 min-1, respectively). The authors conclude that patients receiving isoflurane and enflurane require markedly lower vecuronium infusion rates to achieve 90% neuromuscular blockade than those receiving fentanyl. The enhancement of neuromuscular blockade by isoflurane and enflurane represents a change in the pharmacodynamics of vecuronium-induced neuromuscular blockade, rather than a change in pharmacokinetics.