Diastolic filling of the left ventricle, as assessed by transesophageal pulsed Doppler echocardiography during and in the early phase following coronary artery bypass grafting, was investigated in nine patients without valvular disease or left ventricular hypertrophy. The ratio between the maximal heights of the early diastolic flow-velocity peak and the late (atrial) diastolic flow-velocity peak, the E:A ratio, and also the deceleration time of the early peak were calculated as indices of left ventricular filling. The E:A ratio decreased from 1.01 +/- 0.06 after induction of anesthesia to 0.46 +/- 0.06 on arrival in the intensive care unit (ICU). The E:A ratio then increased and reached 0.87 +/- 0.08 after 6 hours in the ICU. This increase in E:A ratio in the ICU was due to an increasing peak velocity of the E wave. The deceleration time decreased after surgery and increased from 111 +/- 25 to 145 +/- 12 milliseconds in the ICU. Pulsed Doppler indices of diastolic filling are known to be altered by changes in hemodynamic parameters, to be load-dependent, and to vary with heart rate. In the ICU, pulmonary capillary wedge pressure remained unchanged, heart rate decreased by approximately 12%, and systemic vascular resistance decreased by approximately 40%. The changes in hemodynamic parameters could have affected the E:A ratio, but it is unlikely that they could explain the marked increase in the E:A ratio that occurred in the ICU. The results, therefore, imply the presence of impaired diastolic filling immediately after cardiopulmonary bypass with gradual, but not complete, recovery during the first 6 hours in the ICU.