In order to evaluate the usefulness of transcranial Doppler (TCD) in carotid artery surgery, 90 patients who underwent carotid endarterectomy were studied. Transcranial Doppler was performed preoperatively with digital compression of the carotid artery and intraoperatively at the time of cross-clamping, together with routine EEG monitoring. The percentage residual middle cerebral artery velocity (%MCAV) during carotid artery compression or cross-clamping was considered. The intraoperative %MCAV at cross-clamping was compared with the values obtained in the preoperative examination during the compression test and with the EEG data; the EEG records were blindly reviewed by a neurologist and divided into three groups: group A, normal; group B, moderate anomalies and group C, major anomalies. There were no significant differences between preoperative %MCAV (compression test) and intraoperative %MCAV (cross-clamping) by separate analysis of the data related to systolic, diastolic and mean velocity. The systolic, diastolic, mean %MCAV and the Gosling index after cross-clamping were then compared with the EEG data: no significant differences were seen between groups A and B; on the other hand, in all patients with major EEG anomalies (group C) the %MCAV at TCD was reduced to 0. In conclusion, preoperative TCD associated with carotid artery compression appears a safe and useful way of identifying patients at risk of cerebral ischaemia during carotid artery cross-clamping. Patients with no residual flow at intraoperative TCD need to be considered as presenting a very high risk of ischaemia and in this group the use of a shunt is mandatory.