Stroke volume and cardiac output (CO) can be determined noninvasively by means of the pulsed Doppler technique to measure blood flow velocities in specified regions of the heart or neighboring great vessels along with 2D-echocardiographic imaging to measure the diameter of vessels or valve orifices. Disadvantages of the transthoracic approach, such as precordial inaccessibility and instability of the probe position, have prevented the continuous application of pulsed Doppler echocardiography during surgery. Recently, we presented a new technique using the transesophageal approach with combined pulsed Doppler measurements and 2D-echocardiographic imaging. This study was designed to assess the feasibility of transesophageal pulsed Doppler echocardiography (TDE) for CO measurements during surgery and to test the method for accuracy against the thermodilution technique (TD) as well as evaluate its ability to track dynamic CO changes during general anesthesia. Transmitral and pulmonary artery flow analysis using TDE was performed in 35 adult patients undergoing a variety of surgical procedures under general anesthesia. For the transesophageal approach we used the prototype of a new 5-MHz phased array transducer with 64 elements fixed at the distal end of a 9 mm gastroscope. The mitral valve flow methods combined the velocity of transmitral flow at the mitral anulus with the cross-sectional area of the anulus calculated from its diameter at middiastole, while the pulmonary flow method combined the velocity of pulmonary artery flow with the cross-sectional area of the vessel calculated from its diameter during early systole. High-resolution 2D-echocardiograms of the mitral valve allowed accurate diameter measurements of the mitral valve orifice in all patients. A fixed esophageal transducer position behind the left atrium enabled continuous transmitral Doppler recordings of invariably high quality to be made. Regression analysis of TDE-CO vs. TD-CO for 50 measurements in 27 patients yielded a good correlation (r = 0.95, y = 0.95x + 0.42, SEE = 0.34 l/min). Use of halothane in 8 further patients resulted in a 21.0 +/- 5.9% and 37.3 +/- 11.7% decrease of TDE-CO at 1.0 MAC and 1.5 MAC, respectively. Transesophageal images adequate to determine the cross-sectional area of the pulmonary artery could be obtained in 16 of 27 (59.3%) patients. CO determined by the TDE pulmonary flow method (28 measurements in 16 patients) correlated with the TD-CO, with an r value of 0.91 and SEE 0.49 l/min.(ABSTRACT TRUNCATED AT 400 WORDS)