The multiple diameter method previously described and validated for transthoracic echocardiography (TTE) determines ejection fraction (EF) by use of the average of several left ventricular (LV) diameters from multiple views measured at the base, midthird, and distal third of the LV combined with an estimate of the shortening fraction of the long axis (delta L). This method may be ideal for transesophageal echocardiography (TEE) because it does not require tracing of the endocardial contour or volume determinations. Accordingly, EF was calculated with the multiple diameter method in 20 patients in whom TTE and TEE were performed within 1 hour of each other. EF by TTE averaged 49% +/- 20% and ranged from 14% to 80%. The multiple diameter method was modified for TEE as follows: (1) three diameters were taken from the four-chamber view (base, mid-LV, and distal LV) and four from the transgastric view (approximately at 45 degrees from each other), (2) because the LV apex is not well seen by TEE, delta L was estimated from the descent of the mitral anulus towards the apex as 0.15, 0.10, 0.05, or 0 for a descent of > or = 10, 6 to 9, 3 to 5 or < or = 2 mm, respectively. EF by TEE averaged 48% +/- 21% and correlated very well with EF by TTE (r = 0.98; y = 1.03x-2.7). The diameter method was tested prospectively in 30 patients undergoing coronary artery bypass surgery. TEE and TTE were performed within 5 minutes of each other with the patients asleep before initiation of surgery.(ABSTRACT TRUNCATED AT 250 WORDS)