Arteriography and carotid duplex imaging accurately quantify the degree of stenosis caused by carotid atheroma. Since arteriography is inconsistent in identifying carotid ulceration, and controversy exists regarding the diagnostic accuracy of carotid B-mode imaging, a prospective study was performed comparing the two techniques to 126 carotid endarterectomy specimens. Sixty percent (76/126) of specimens contained ulcers. The diagnostic sensitivity for B-mode imaging and arteriography was 47% (36/76) and 53% (40/76), respectively (p = NS). Importantly, the degree of stenosis caused by the plaque significantly affected diagnostic sensitivities. B-mode sensitivity was 77% (10/13) in plaques less than or equal to 50% and 41% (26/63) for plaques greater than 50% (p = 0.03). Arteriography likewise detected 77% (10/13) of ulcers in plaques less than or equal to 50% stenosis and 48% (30/63) in plaques with greater than 50% stenosis (p = 0.07). In patients with focal symptoms, 100% (10/10) of plaques with less than or equal to 50% stenosis contained ulcers, whereas in plaques with greater than 50% stenosis 63% (36/57) contained ulcers (p = 0.02). These data indicate that the diagnostic sensitivity for carotid ulceration is not significantly different between B-mode carotid imaging and arteriography. Since most B-mode errors occur with high-grade stenoses this short-coming is unlikely to adversely affect patient care. Previous studies investigating the ability of carotid B-mode imaging to detect ulceration failed to address quantitative aspects of the carotid plaque. These data appear to resolve the previously existing controversy.