In the diagnosis of extracranial carotid arterial disease, quantitative measurements from the continuous wave (CW) Doppler spectrum have the potential for detecting stenoses and occlusions. The measurement of maximum peak Doppler frequency at the site of stenosis has been shown to detect severe, but not minor or moderate, stenoses. Diagnosis of minor or moderate stenoses may be possible by assessing the degree of flow disturbance beyond the stenosis. Such flow disturbances cause the Doppler spectrum at peak systole to be broadened, and it has been suggested that the measurement of spectral broadening may be of diagnostic value. This paper describes the results of an in vitro study aimed at determining the hemodynamic factors that influence the severity of the Doppler spectral broadening. The spectral broadening index (SBI) at peak systole, defined as SBI = 1 − F mean / F max , was used to quantify the instantaneous spectrum. In a pulsatile flow in vitro model that produced spectral waveforms virtually identical to those recorded in the human carotid, we observed a direct linear relationship between SBI and the severity of stenosis, at least for those stenoses having greater than 40% cross-sectional area ( R = 0.82 to 0.93). The SBI was found to be maximum when recorded immediately beyond the stenosis and returned to normal 4–5 cm downstream from the stenosis. The SBI was higher for nonsymmetrically shaped stenoses than for symmetrical stenoses for lesions greater than 60%, but not for stenoses less than 60%. In this model, the SBI recorded from both normal or abnormal waveforms was not affected by the flow rate. Furthermore, the measurements of SBI in diastole (i.e. 100 msec after the peak) were not related to the severity of the stenosis, whereas the measurements of SBI at peak systole were. We conclude that SBI has potential clinical diagnostic value, since it is related to the severity of the stenosis. Although it is also affected by the shape of the stenosis and by the recording site in relationship to the stenosis (at least for stenoses greater than 60%), as discussed in the paper, we do not feel that these factors will limit its clinical diagnostic value. The diagnostic accuracy of SBI should be assessed in a prospective clinical study.