Physicians have been increasingly relying on noninvasive imaging methods to grade carotid stenosis. The accuracy of Doppler ultrasound (DUS) and CT angiography (CTA) versus intra-arterial angiography (IA) has been assessed in many studies and at least two meta-analyses. Here, we performed a systematic review of studies that compared DUS to CTA. In a PubMed review of the literature from 2000 to 2009, we found 12 studies that compared DUS and CTA-based grading of carotid stenosis. Only 4 of them included at least 20 patients and provided data to classify the diseased arteries into the following categories: mild, moderate or severe NASCET stenosis or occlusion. We extracted 431 arteries from 244 patients (range per study: 48-164). It was not possible to distinguish symptomatic from asymptomatic stenoses. Nearly half arteries had severe stenosis (46% based on DUS and 43% based on CTA). The weighted kappa was 0.85 (95% CI 0.76-0.94), and the accuracy was 0.78. When the arteries were classified into medical and potentially surgical groups, the kappa was 0.76 (95% CI 0.70-0.83), and the accuracy was 0.89. Overall, 17% of the stenoses classified as medical based on DUS were reclassified as surgical based on CTA and 14% of the stenoses classified as medical based on CTA were reclassified as surgical based on DUS. The sparse available data comparing DUS and CTA suggest that the grading of a carotid stenosis as medical or potentially surgical remains uncertain in a relatively high proportion of patients.