An ultrasonic sector scanner and pulsed Doppler flowmeter were combined to measure portal venous velocity in 19 patients with portal hypertension and compared with cineangiographic mapping of a droplet of oil released into the portal vein. In 12 patients, measurements were made before and after pitressin [7] or ranitidine [5]. With the Doppler method, maximum basal portal venous velocity was 17.0 +/- 3.9 cm/sec., while average cineangiographic velocity was 8.5 +/- 2.7 cm/sec., a significant difference (p less than 0.001). After pitressin, velocity decreased to 8.3 +/- 2.7 cm/sec. with the Doppler method and 3.6 +/- 1.0 cm/sec. with cineangiography. Ranitidine did not produce an appreciable change. Doppler and cineangiographic velocity measurements exhibited significant correlation over a wide range of values (r = 0.960). Thus the pulsed Doppler method may give accurate values of portal venous velocity if they are corrected to cineangiographic values. Flow can be calculated from velocity and the cross-sectional area of the portal vein as measured on the sonogram. The Doppler method is simple and noninvasive and is particularly useful in studying changes in portal hemodynamics.