A combined pharmacokinetic and pharmacodynamic model of mcthohcxital was used to establish and evaluate feedback control of methohexital delivery during total intravenous anesthesia with fentanyl in 11 surgical patients. The median frequency of the EEG power spectrum served as the pharmacodynamic variable constituting feedback. Based on previous investigations a median frequency from 2–3 Hz was chosen as the desired EEG set point. In addition to methohexital, patients were given a 10-min loading infusion of 0.5 mg of fentanyl followed by a constant-rate infusion of 0.22 mg/h. In agreement with an earlier similar study in volunteers given only methohexital and aiming at the same set point, identical distribution of EEG power was achieved in the current study. The decrease of median EEG frequency to 2–3 Hz was primarily induced by an increase in fractional power in the 0.5–2-Hz frequency band to 46 ± 4%. The average requirement of methohexital during the first 2 h was 675 ± 250 mg. The authors conclude that model-based feedback control of intravenous methohexital delivery can help establish and quantitate methohexital requirements during total intravenous anesthesia with fentanyl.