To evaluate possible factors affecting the pharmacokinetics of thiopentone during cardiopulmonary bypass (CPB), the present study was undertaken in patients scheduled for coronary artery bypass grafting and with in vitro experiments. The effects of nonpulsatile and pulsatile flow during CPB on the distribution and elimination of thiopentone were compared in 30 patients anaesthetized with fentanyl. The initial rapid phases of distribution of thiopentone were studied in 17 patients undergoing a nonpulsatile or pulsatile perfusion, to whom thiopentone 6 mg/kg was given as a rapid intravenous bolus during CPB. In order to study later distribution and early elimination of thiopentone, 13 patients perfused with a nonpulsatile or pulsatile flow received 6 mg/kg of the drug as a 15-min intravenous infusion before CPB. No differences in the pharmacokinetic parameters characterizing distribution and elimination of thiopentone were found between the patients undergoing nonpulsatile or pulsatile perfusion. As measured in 10 of the patients receiving the drug before the institution of CPB, no difference in plasma thiopentone level was observed in blood samples drawn simultaneously from a radial arterial cannula and a pulmonary artery catheter before, during and after CPB. This suggests that thiopentone is not sequestered in lungs during CPB. In vitro binding of thiopentone to the CPB equipment was studied in 6 experiments using a closed circuit. After a 60-min circulation time, only 50% of the predicted thiopentone level was recovered from the perfusate. It is concluded that replacing a nonpulsatile perfusion with a pulsatile one has no effect on the distribution and elimination of thiopentone in patients undergoing CPB. During CPB, thiopentone is sequestered in the extracorporeal circuit but not in the lungs.