One hundred patients were randomly allocated to receive saline or amrinone, 0.75 mg.kg-1, ten minutes before separation from cardiopulmonary bypass (CPB) after elective coronary artery bypass grafting, in order to determine the effects of this agent on haemodynamic variables and O2 utilization. Anaesthesia and CPB were managed in a standard fashion. Before induction of anaesthesia, at pericardiotomy, then at 1, 10, 20 and 30 min after CPB, haemodynamic profiles, haematocrit, and O2 saturation of arterial and mixed venous blood were measured. Incremental doses of ephedrine or phenylephrine, or an infusion of norepinephrine with phentolamine were administered when required. The groups were demographically similar and surgical variables were also similar. Haemodynamic measurements were similar between groups at all times; however, a higher dose of phenylephrine was given immediately before weaning from CPB in the amrinone group, and more patients in this group received phenylephrine in the first 30 min after CPB. Mixed venous saturation (SvO2) was higher in the amrinone patients at all times after CPB, leading to lower calculated oxygen consumption (VO2) (P less than 0.05). Insufficient dosage may explain the lack of haemodynamic effect, while possible reasons for the higher SvO2 and lower VO2 are either reduced whole body VO2 or peripheral shunting.