Objective: To determine the pharmacokinetics of sufentanil in patients undergoing coronary artery bypass graft surgery. Design: Prospective, multigroup study. Setting: University-affiliated hospital. Participants: Patients with good left ventricular function undergoing elective surgery (n = 103). Interventions: Sufentanil was administered by target-controlled infusion, with target effect-site concentrations ranging from 0.4 to 4.5 ng/mL. Isoflurane was administered as required to maintain stable hemodynamics. Sufentanil pharmacokinetics were determined by population modeling. The potential effects of gender, weight, different premedications (lorazepam, morphine-scopolamine, or clonidine), and coinduction with propofol on sufentanil pharmacokinetics were explored. Measurements and Main Results: The first model determined was a simple 3-compartment model, without any covariates, which had these parameters: V-1 = 5.7 L, V-2 = 18.1 L, V-3 = 225 L, Cl-1 = 0.69 L/min, Cl-2 = 3.1 L/min, and Cl-3 = 1.4 L/min. The overall predictive ability during the entire pre-cardiopulmonary bypass period of this model was excellent, with virtually no bias (median prediction error, -0.4%) and good precision (median absolute prediction error, 18.4%). More complex models with the various premedications used or coinduction with propofol as covariates did not improve the predictive accuracy or precision compared with the simple 3-compartment model. Similarly, including either gender or weight as a covariate did not improve predictive ability. Conclusion: The authors have determined a pharmacokinetic model for sufentanil that can be used to maintain desired target concentrations of sufentanil before cardiopulmonary bypass, with a high degree of accuracy and acceptable variability. Concomitantly administered medications (lorazepam, morphine-scopolamine, clonidine, or propofol) do not appear to have any clinically important effects on distribution-phase sufentanil pharmaco kinetics. Copyright (C) 2001 by W.B. Saunders Company.