PURPOSE: This study was undertaken to prospectively evaluate the diagnostic performance of colour Doppler ultrasonography (CDUS), first-pass (FP) and steady-state (SS) contrast-enhanced magnetic resonance angiography (MRA) and computed tomography angiography (CTA) of the carotid arteries using digital subtraction angiography (DSA) as the reference standard. MATERIALS AND METHODS: A total of 170 patients with previous cerebrovascular events and suspected carotid artery stenoses underwent CDUS, blood-pool MRA, CTA and DSA. Accuracy, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated for CDUS, FP MRA, SS MRA and CTA. The McNemar and Wilcoxon tests and receiver operating characteristic (ROC) curve analysis were used to determine significant differences (p<0.05) between the diagnostic performances of the four modalities, and the degree of stenosis was compared using linear regression. RESULTS: A total of 336 carotid bifurcations were studied. The area under the curve (AUC) for degree of stenosis was: CDUS 0.85±0.02, FP MRA 0.982±0.005, SS MRA 0.994±0.002 and CTA 0.997±0.001. AUC analysis showed no statistically significant difference between CTA and MRA (p=0.0174) and a statistically significant difference between CDUS and the other techniques (p<0.001). Plaque morphology analysis showed no significant difference between CTA and SS MRA; a significant difference was seen between CTA and SS MRA versus FP MRA (p=0.04) and CDUS (p=0.038). Plaque ulceration analysis showed a statistically significant difference between MRA and CTA (0.04< p<0.046) versus CDUS (p=0.019). CONCLUSIONS: CTA is the most accurate technique for evaluating carotid stenoses, with a slightly better performance than MRA (97% vs. 95% for SS MRA and 92% for FP MRA) and a greater accuracy than CDUS (97% vs. 76%). Blood-pool contrast-enhanced SS sequences offer improved evaluation of degree of stenosis and plaque morphology with accuracy substantially identical to CTA.