Several efforts have been made to improve the technique of cardiopulmonary bypass (CPB), including the use of pulsatile flow and the modification of cannulation technique. The present study focused upon extravascular lung water (EVLW) in 60 aortocoronary bypass patients subjected to four different perfusion techniques during CPB: group 1 (n = 15): non-pulsatile flow and standard cannulation; group 2 (n = 15); pulsatile flow and standard cannulation; group 3 (n = 15): nonpulsatile flow and monoatrial cannulation (i.e., always "partial" bypass during CPB); group 4 (n = 15): pulsatile flow and monoatrial cannulation. EVLW content was measured using the double-indicator dilution technique with indocyanine green; in addition, various hemodynamic and laboratory variables were measured. Lung water content rose above normal values (mean: 5.79 +/- 0.33 ml/kg) only in the groups submitted to the standard cannulation technique, irrespective of whether the perfusion flow was pulsatile or not (group 1: + 27.4%; group 2: + 25.5%). Pulmonary gas exchange, too, was compromised only in these patients (PaO2 in group 1 -19% and in group 2 -17%; Qs/Qt in group 1 + 36 rel. % and in group 2 + 29 rel. %), whereas all patients with monoatrial cannulation and partial bypass throughout the CPB period showed no rise in EVLW content or Qs/Qt and no drop in PaO2. From the results of this study we conclude that pulsatile perfusion during open heart surgical procedure has no advantages in regard to lung water content. Monoatrial cannulation with partial bypass at all times during CPB, however, seems to be beneficial, probably owing to the maintenance of pulmonary circulation during the bypass period.