Hemodynamic effects of pressure support ventilation (PSV) were assessed in 33 patients, after cardiac surgery. The patients, selected for their stable left ventricular and respiratory functions, underwent uncomplicated coronary artery bypass grafting. They all underwent the same anesthetic protocol, and an invasive hemodynamic monitoring. Eight to ten hours postoperatively, while all patients were fully awake, normothermic and hemodynamically stable, controlled ventilation (CV) was replaced by 3 levels of PSV (20, 10, 5 cm H2O). These levels were applied consecutively for 20 minutes each, before extubation. Hemodynamic and gas exchange data were recorded on CV, on each level of PSV, and on spontaneous breathing. The results were analyzed using ANOVA and Bonfferoni methods. No statistical significance could be noted between the five modes of ventilation as to hemodynamic parameters, arterial and mixed venous blood gases, and oxygen consumption (VO2). The only 2 parameters that reached statistical significance were central venous pressure and respiratory rate. Our study demonstrates that patients with stable cardiovascular and respiratory function can adapt to different levels of PSV without hemodynamic modifications or an increase in their VO2.