The aim of this study is to compare measurement of stump pressure (SP) and somatosensory evoked potentials (SSEP) made during carotid surgery as criteria upon which to base the decision whether or not to use a shunt. We included 288 patients who underwent for carotid surgery under general anaesthesia. We performed 247 endarterectomies with patch closure (85.7%), 25 carotid transsection with reimplantation (8.7%), and 16 carotid bypasses (5.6%). SSEP monitoring showed no modification in 225/288 patients (78.1%), moderate modification in 32/288 patients (11.1%), and severe modification in 31/288 patients (10.8%). Shunt was used if there was moderate or severe SSEP modification in response to carotid clamping, which represents 63 patients in our series. A shunt was used in 47/288 patients (16.3%). In 16 patients, despite SSEP modifications, the shunt was not used because these SSEP modifications occurred only in the last minutes of the procedure just before off clamping the carotid. The mean SP for all patients was 51 mm Hg. In the shunted patients, the mean SP was 33 mm Hg. Variation of SP was correlated with the SSEP modifications. There was just one perioperative stroke in this series (1/288 = 0.3%). We concluded that the threshold of SP below which shunting is indicated in our study was 44 mm Hg with 81% sensibility and 68% specificity.