New criteria for the interpretation of duplex scanning waveforms were compared with those devised by Blackshear and Strandness and revealed an improved grading of internal carotid stenosis in early lesions. The ratio of peak frequency minus first zero slope frequency divided by the peak frequency (a-b/a ratio) was found to be of no value in the discrimination between normal and any grade of carotid stenosis. The new criteria depend upon the to dental presence of spectral broadening on the upstroke of systole to differentiate between normal and abnormal, spectral broadening throughout the systolic phase above the end diastolic frequency to determine greater than 15% stenosis, peak frequency greater than 4kHz to indicate greater than 50% stenosis and the absence of a Doppler signal to indicate occlusion. Comparison with two plane selective carotid angiography revealed a Kappa statistic of 0.8 +/- 0.02 using the new criteria compared with 0.65 +/- 0.03 using the established criteria of Doppler waveform analysis.