Objective:To determine if the cardiac output (CO) measured by transesophageal echocardiography (TEE) through the left ventricular outflow tract (LVOT) is consistent with that measured by pulmonary artery catheter (PAC). The correlation between left ventricular ejection fraction (LVEF),left ventricular fractional area change (LVFAC) and right ventricular ejection fraction (RVEF) was analyzed. Methods:Twelve patients with ASA Ⅱ~Ⅲ(NYHA Ⅱ or Ⅲ),aged 18~70 years,weighing 46~72 kg and undergoing coronary artery bypass grafting were studied. Anesthesia induction and intraoperative maintenance were performed by intravenous anesthesia. After tracheal intubation,the Swan-Ganz catheter and TEE probe were placed. The data were measured and recorded after tracheal intubation (T0),15 (T1),30 (T2) and 60 (T2) min after termination of cardiopulmonary bypass (CPB) or finished vascular anastomosis in off-pump CABG. Statistical analysis was performed by Bland-Altman plot and Pearson correlation coefficient method. Results:COPAC was (4.82 ± 1.32) L/min and COLVOT was (4.57 ± 1.30)L/min. COLVOT was highly correlated with COPAC (r = 0.655,P < 0.001). The bias between COPAC and COLVOT was 0.28 L/min (95%CI:-0.04~0.60 L/min),and limits of agreement were -1.90~2.46 L/min. LVEF was positively correlated with LVFAC (r = 0.662,P < 0.001),while it was not significantly correlated with RVEF (r = -0.218,P > 0.001). The bias between LVEF and LVFAC was 15.36% (95%CI:12.46%~18.29%),and limits of agreement was 4.68%~35.43%,while that between LVEF and RVEF was 34.40% (95%CI:29.69%~39.10%),and limits of agreement was -2.01%~66.78%. Conclusion:The results showed that CO measured by TEE through the LVOT was significantly correlated but poorly consistent with that measured by Swan-Ganz PAC. Both methods for CO measurement cannot replace each other. Moreover,there was an obviously bias but a good correlation between LVFAC and LVFAC.