BACKGROUND Transesophageal echocardiography (TEE) has been regarded as a standard diagnostic method for assessment of cardiac output in aortic dissection, pulmonary embolism, atrial and ventricular septal defects, and acquired valvular defects. It is also a standard method of assessment of the adequacy of their surgical repair. It allows for on-line estimation of ejection fractions and stroke volumes, and calculation of cardiac output. In this prospective study, we compared the cardiac outputs estimated by aortic continuous-wave Doppler, and pulse wave Doppler via transgastric long-axis imaging plane TEE, with results obtained by thermodilution. METHODS Cardiac output was determined in thirty adult patients, of both sexes, with good left ventricular function, scheduled for off pump coronary artery bypass grafting. Aortic valve area was plotted from the transverse short-axis view of the valve assuming a triangular shape for the valve orifice. RESULTS Mean cardiac output measured by thermodilution [CO(S-G)] was 4.59 +/- 2.5 L min(-1), compared to 4.49 +/- 1.14 L min(-1) obtained from contour wave [CO(CW)] and 4.57 +/- 1.29 L min(-1) from continuous-wave [CO(PW)] analysis. The correlation coefficient (r) between CO(S-G) and CO(CW) was 0.939, (p < 0.001).The correlation coefficient (r) between CO(S-G) and CO(PW) was 0.912 (p < 0.001). CONCLUSIONS Results of cardiac output measurements, obtained from all three methods, were comparable.