Twenty-eight patients undergoing elective coronary artery bypass surgery were allocated randomly to receive either propofol 2 mg kg-1 or an equivalent volume of its vehicle during cardiopulmonary bypass with constant pump flow. Peripheral vascular resistance (PVR) was calculated from perfusion pressure and pump flow. After propofol, PVR decreased from 1767 (SD 415) dyn s cm-5 to a minimum of 1263 (283) dyn s cm-5 at 2 min, and remained significantly less than the control value until 12.5 min after administration of propofol. In the group given the vehicle, PVR did not change significantly. In a second study in 10 patients, venous blood samples were withdrawn before and 2, 4, 6, 8, 10, 20 and 30 min after injection of propofol 2 mg kg-1 during cardiopulmonary bypass, for measurement of blood concentrations of propofol. Concentrations were greater than predicted by a computer simulation based on published pharmacokinetic data. The decrease in PVR may be an important factor in the hypotension caused by propofol during induction of anaesthesia.