With the single Doppler spectrum analysis, one can appreciate the degree of the carotid stenosis according to the importance of the haemodynamic disturbances induced by the stenosis. The purpose of our study is to show the possibilities and the limitations of this method. Spectrum disturbances were classified in 5 grades, each of them being related to the importance of the stenosis. The degree of stenosis has been evaluated by the C.W. Doppler spectrum analysis, the angiographies and the anatomical study of the endarteriectomies. We considered that we had a perfect concordance between the results of the different methods when the degree of stenosis measured on the angiographies or on the endarteriectomies was compatible with the spectrum analysis classification: grade I: stenosis inferior to 40% (in area), 23% (in diameter); grade II: stenosis ranging between 40 and 60% in area (23 and 40 in diameter); grade III: stenosis of 60 to 75% in area (40 to 50% in diameter) and of particular shape (extended plaque); grade IV: stenosis ranging between 60 and 90% in area (40 to 70% in diameter); grade V: stenosis higher than 90% in area (70% in diameter). The confrontation of spectrum analysis and angiographic date concerns 58 bifurcations. We got a perfect correlation in 93% of the cases. The confrontation of the spectrum analysis method and the anatomical study of the endarteriectomies concerns 38 bifurcations. We got a perfect correlation in 92% of the cases. The appreciation of the carotid stenosis degree is now performed in routine at the Hospital. For some patients, the endarteriectomy has been decided from the clinical and the spectrum analysis data, and an electro-encephalogram with compression. However these date are generally completed with an angiography with venous punction.