OBJECTIVE Transcranial Doppler ultrasonography (TCD) is the standard technique for monitoring emboli in the cerebral circulation. Embolic signals have been detected with the use of this technique in most patients undergoing coronary artery bypass surgery. We previously reported that the majority of emboli are detected after release of aortic cross-clamps and partial occlusion clamps. In this study we compare the intraoperative use of TCD with transesophageal echocardiography (TEE) to monitor cerebral emboli. METHODS We simultaneously monitored 20 patients undergoing coronary bypass surgery with TCD and TEE. All patients also underwent routine TEE examination of the aorta. RESULTS Embolic signals were detected in all patients by both techniques. Mean total number of emboli was 535 +/- 109 by TEE compared with 133 +/- 28 by TCD. We found correlation between numbers of emboli detected by the two techniques at clamp placement and release (r = .65, P = .002). Clamp placement and release accounted for 84% of all emboli by TEE and 83% by TCD. By TEE, large, highly echogenic particles were detected after clamp release compared with small, barely echodense particles at the onset of bypass. No such distinction was apparent by TCD. We found correlation between severity of aortic atheroma and both TEE- (P = .003) and TCD-detected (P = .009) emboli. CONCLUSIONS TEE and TCD can both be used to continuously monitor emboli during coronary artery bypass surgery. However, TEE is invasive and justified only if it is being performed for intraoperative assessment of aortic atheromatosis or cardiac function.