OBJECTIVE The purpose of this study was to determine whether noninvasive evaluation with duplex ultrasonography and magnetic resonance angiography of patients with carotid artery stenosis can replace contrast angiography at our institution. METHODS This study consisted of a retrospective chart review of 40 patients (74 carotid arteries) in combination with a blinded reanalysis of original data. Contrast angiography was compared with duplex ultrasonography and magnetic resonance angiography. The overall diagnostic accuracy of duplex ultrasonography and magnetic resonance angiography was determined individually and concordantly in patients being evaluated for carotid artery stenosis. RESULTS The overall sensitivity of duplex ultrasonography was 88.5%, and the specificity was 91.7% (Spearman correlation coefficient = 0.8456; p < 0.001). For magnetic resonance angiography the sensitivity was 92.3%, and the specificity was 97.9% (Spearman correlation coefficient = 0.9086; p < 0.001). In the presence of concordance, the noninvasive studies exhibited a sensitivity of 100%, (correlation coefficient = 0.9661; kappa value = 0.9655). No occlusions or severe lesions were missed by both studies. In only one vessel (1.52%) was a false-positive concordance noted. CONCLUSIONS Carotid endarterectomy may be undertaken with a high degree of confidence that the operation will be appropriate if the noninvasive evaluations are concordant. In the absence of concordance of the noninvasive studies, contrast angiography should be considered.