OBJECTIVE The degree of stenosis measured by Doppler ultrasonography in patients with contralateral occlusion of the internal carotid artery (ICA) is assumed to be overestimated. We propose to measure the extent to which this phenomenon affects the capacity of Doppler ultrasonography to classify stenoses. METHODS A retrospective study of 47 patients was conducted using Doppler ultrasonography and digital subtraction angiography; all subjects had confirmed unilateral occlusion of the ICA and stenoses of the contralateral ICA. Curves were traced plotting the degree of stenosis against the peak systolic velocity (PSV) and its ratio in the ICA and in the common carotid artery (VICA/VCCA). The curves were compared with their equivalents with no contralateral occlusion found in the literature. Later, the cases of stenosis were classified into groups of less than 50, 50-69 and more than 70%, and the main statistical values were calculated. RESULTS The PSV in the stenoses between 40-50% presents more than 1 standard deviation (SD) above the mean. There were no significant deviations in other degrees of stenosis (p > 0.4). The VICA/VCCA showed a similar parallelism, but with deviations lower than 1 SD (p = 0.56). Seventeen per cent of the stenoses were over-classified, and this conditioned a sensitivity of 84, 71 and 100%, and a specificity of 100, 94 and 88% for the groups of less than 50, 50-69 and over 70%, respectively. The VICA/VCCA over-classified 41% of the stenoses; sensitivity was seen to be 56, 43 and 100% and specificity was 90, 64 and 87%. Diagnostic accuracy of the PSV and VICA/VCCA stands at 83 and 57%, respectively. CONCLUSIONS Contralateral occlusion leads to over-classification of the PSV. There is a tendency to over-classify, although this does not affect the overall diagnostic accuracy. The VICA/VCCA does not offer greater diagnostic accuracy in classifying stenoses with contralateral occlusion.