Clinical management of carotid disease is primarily based on results of national trials (ACAS and NASCET) that used the distal internal carotid artery diameter as a reference. However, commonly accepted ultrasound (US) criteria for carotid stenosis were derived from the correlation of velocity measurements with angiographic bulb diameter reductions (BDR). This study was undertaken to compare the degree of carotid stenosis determined by conventional velocity criteria to the degree of stenosis measured by B-mode (gray scale) diameter at both the carotid bulb and at the distal internal carotid artery, and, second, to evaluate US imaging to derive distal diameter reductions (DDR) noninvasively. During a 3-month period patients referred for carotid US were prospectively analyzed for standard velocity criteria and plaque morphology. Minimum carotid diameter was measured by longitudinal and transverse B-mode measurements and compared to carotid bulb diameter and internal carotid diameter distal to all disease. B-mode diameter reductions were compared to the degree of stenosis determined by velocity criteria and to patient symptoms and the decision for carotid endarterectomy. In total, 131 carotid arteries in 74 patients were evaluated. Based on the University of Washington velocity criteria, lesions were classified as grade I (n = 61, 46%), IIA (n = 58, 44%), IIB (n = 7, 5%), or III (n = 5, 4%). BDR measured by B-mode predicted the grade of disease based on velocity criteria (p < 0.001) with an overall accuracy of 95%. With use of the B-mode for DDR (NASCET style), 18 patients exceeded the 60% threshold for surgical intervention. Of these, only 3 patients were symptomatic and were operated on. An additional 3 operated-on patients had an asymptomatic grade III stenosis, our usual threshold for intervention. Twelve additional patients were appropriate for surgical intervention by B-mode but were not treated based on conventional velocity criteria alone. Bulb diameter reduction by B-mode imaging correlates strongly with diameter reduction determined by velocity criteria, and independently predicts the grade of carotid disease. With this in mind, the accuracy of B-mode imaging may be extended to the measurement of carotid stenosis based on DDR. By B-mode criteria, many patients appropriate for intervention were not offered treatment based on conventional velocity criteria. Modern B-mode imaging provides a noninvasive method to obtain 'arteriographic equivalent'' measurements and should be added as a routine to carotid ultrasound interrogation.