Intraoperative real-time B-mode ultrasonography was used to evaluate the technical results of 155 carotid endarterectomies in 143 patients. Technical defects created as a result of the endarterectomy were detected in 43 of the 155 endarterectomies (27.7%) and included intimal flaps (73% of defects); strictures (18%); and arterial kinks, residual plaque, and intraluminal thrombi (9% collectively). Eleven of the 43 endarterectomy sites (7% of all endarterectomies) were reentered to correct a defect; none of these patients had neurologic deficits, which suggests that reentering an endarterectomy and correcting a defect does not, in and of itself, lead to a higher incidence of stroke. The incidence of stroke in patients with normal results of intraoperative ultrasonography was 3.8%, whereas the incidence of perioperative stroke in those patients with insignificant and thus uncorrected defects was 3.3%; this suggests that intraoperative ultrasonography is sufficiently sensitive to detect defects that, when left uncorrected, do not lead to a higher than usual incidence of stroke. Because intraoperative ultrasonography is safe and highly sensitive, we believe it is the method of choice for assessing the technical results of carotid endarterectomy.