In order to evaluate the suitability of transcranial Doppler sonography as an intraoperative monitor in carotid surgery, we compared measurements of mean blood flow velocity in the ipsilateral middle cerebral artery with the cortical response of somatosensory evoked potentials in a prospective study of 176 carotid operations. SEP recording was readily feasible during all procedures and by means of SEP loss all patients at risk for critical cross-clamp related cerebral ischemia were reliably identified. In contrast, TCD could not be used for assessment of cerebral hemodynamics in more than 40% of patients. What is more, in high risk patients with intraoperative loss of SEP, TCD could not be performed in 74% of cases. This high rate of failure limits the usefulness of TCD as an intraoperative monitor and detracts from the additional benefit of identifying cerebral embolism and hyperperfusion as potential causes of neurological deficits. In contrast to SEP recording, TCD cannot be recommended as a routine monitor in carotid surgery.