Hemodynamic responses to pipecuronium bromide or doxacurium chloride were compared in patients undergoing valvular heart surgery. Thirty ASA class III-IV patients of either sex, mean age 62 ± 3 years (±SD), weight 70 ± 3 kg, were randomly selected to receive either doxacurium (0.08 mg/kg) or pipecuronium (0.15 mg/kg). Hemodynamic parameters were determined at preinduction, induction, 2 minutes and 6 minutes following administration of the muscle relaxant. Anesthetic induction consisted of midazolam, 0.10 mg/kg, followed by fentanyl, 5 μ/kg. Measured or calculated parameters were as follows: mean arterial pressure, heart rate, cardiac index, mean pulmonary artery pressure, systemic vascular resistance index, central venous pressure, pulmonary capillary wedge pressure, stroke volume index, left and right stroke work indices, and pulmonary vascular resistance index. Awake patients who had been randomly assigned to the pipecuronium group had significantly higher pulmonary capillary wedge pressures (22 ± 2 v15 ± 2 mmHg; P< 0.05) and heart rates (86 ± 3 v64 ± 5 beats/min; P< 0.05) than awake patients in the doxacurium group. Following induction, both wedge pressure and heart rate were not significantly different between the two groups. Compared to hemodynamics at induction, there were no clinically significant changes following administration of pipecuronium or doxacurium.