The authors analyse their experience with carotid stump pressure (CSP) and electro-encephalogram (EEG) monitoring in relation to the incidence of peroperative stroke during internal carotid artery reconstruction. A series of 215 patients is presented, among whom six (2.8%) developed a peroperative stroke. The stroke rate in patients with a CSP below 50 mm Hg (n = 92) was 7% (2/27) without a shunt and 3% (2/62) with the use of a shunt. The respective numbers for patients with a CSP equal to or above 50 mmHg were 1.7% (2/123) and 0% (0/10). The EEG remained normal after cross clamping in 180 cases (84%): the incidence of stroke was 1.5% (2/138) without and 5% (2/42) with a shunt. A shunt was used in 33 of the 35 patients with EEG changes after cross clamping. None of them sustained a stroke in contrast to both patients where despite EEG changes no shunt was used (respective stroke rates 0% and 100%). It is concluded that regarding cerebral function, EEG monitoring provides more accurate information than CSP.