We compared three angiographic methods for grading of carotid stenosis and examined the correlation between angiographic and ultrasound findings. Two observers independently measured 111 carotid stenoses on arteriographic films of 84 patients. The stenoses were graded according to the European Carotid Surgery Trial (ECST), North American Symptomatic Carotid Endarterectomy Trial (NASCET), and Common Carotid (CC) methods. The results obtained by these methods were compared, and the interobserver reproducibility of the measurements was calculated. In addition, all angiographic results were compared to ultrasound findings obtained before angiography. Measurements using the CC method were the most reproducible and those using the NASCET method the least. The NASCET method underestimated the degree of stenosis compared to the other methods. The ECST and CC methods yielded almost identical results (97% agreement). Ultrasound provided an accuracy of 94% compared to ECST and CC methods and 84% compared to the NASCET method. Interobserver reproducibility of angiographic quantification of carotid stenoses was best for the CC and ECST methods and least for the NASCET method. Ultrasound demonstrated better accuracy than the ECST and CC methods.