Nine patients, all NYHA class III and IV, who were to undergo surgery with cardiopulmonary bypass, were divided into two groups. All nine were premedicated with 2 mg flunitrazepam orally 90 min before surgery. Group A consisted in five men, mean age 53 years (extremes: 43 and 73), with no significant difference in weight and body surface area, who were to undergo aortic valve replacement (2), mitral valve replacement (2), and aortic and mitral valve replacement (1). They were given 1.5 mg X kg-1 propofol as a bolus at induction, followed by an infusion of 50 micrograms X kg-1 X min-1 propofol. Group B consisted in four patients, two men and two women, mean age 60 years (extremes: 49 and 76), with no significant difference in weight and body surface area, who were to undergo aortic valve replacement (3) and corrective surgery for left ventricular aneurysm (1). They were given 2 mg X kg-1 propofol as a bolus at induction, followed by an infusion of 100 micrograms X kg-1 X min-1 propofol. All the patients were intubated after 0.1 mg X kg-1 pancuronium; they were given 10 micrograms X kg-1 before intubation and 10 micrograms X kg-1 before skin incision. Repeat doses of fentanyl and pancuronium were given as required. The haemodynamic parameters studied were: heart rate, systolic, diastolic and mean blood pressures, central venous pressure, mean pulmonary pressure, wedge pressure, and cardiac output measured by thermodilution. The following indices were calculated: cardiac index, systolic index, right and left systolic work indices, and pulmonary and systemic vascular resistances.(ABSTRACT TRUNCATED AT 250 WORDS)