Continuous monitoring of cardiac output during liver transplantation is essential to evaluate the patient's haemodynamic tolerance to acute volume variations. The aim of this study was to compare the cardiac output values obtained with a transoesophageal echo-Doppler and those obtained with a continuous thermodilution cardiac output pulmonary artery catheter. Twenty adult patients were prospectively studied during a 5 min hepatic vascular exclusion test performed at the end of the dissection phase. Echo-Doppler and continuous thermodilution cardiac output, mean arterial pressure and end-tidal CO2 were measured before and at the end of the test. Before the test, echo-Doppler cardiac output was 7.0 +/- 2.7 L min(-1) and thermodilution was 9.4 +/- 3.1 L min(-1), (R = 0.85, P < 0.001). The end test values were, respectively, 3.5 +/- 2.7 and 7.8 +/- 3.5 L min(-1) (R = 0.23, P = 0.34). Bland and Altman analysis showed a bias of -2.2 before the test, which increased to -4.4 at the end of the test. Mean arterial pressure decreased from 85.5 +/- 15 to 66.8 +/- 16 mmHg, end-tidal CO2 from 31.4 +/- 2.3 to 23.8 +/- 2.7 mmHg. Echo-Doppler cardiac output values are different from those measured by thermodilution cardiac output in these patients. Echo-Doppler cardiac output monitoring seems to detect the output changes, which can occur during acute haemodynamic changes more rapidly than thermodilution cardiac output in the course of liver transplantation.