Although the ultrasonic pulsatility index (PI) is used as a test of arterial inflow, the relationship between proximal arterial stenosis and PI has not been fully evaluated. In a series of experiments on dogs, over 200 measurements of pressure, flow (using an indwelling electromagnetic flowmeter), and PI were made distal to implanted arterial stenoses of varying length (0.5-9 mm) and cross-sectional area (36-94% reduction in area). It was shown that, whereas the reduction of PI correlates broadly with stenosis severity, the scatter of results is wide for all but the tightest stenoses. This scatter is due at least in part to the variability of the vascular bed distal to the site of measurement. Circ Res 46: 470-475, 1980