In order to evaluate the behavior of the portal vein cross-sectional area during changes in portal flow, two groups of subjects were analyzed in two blinded cross-over studies using echo-Doppler flowmetry. The first group (I) consisted of 21 patients with cirrhosis and 16 controls. They received a standardized meal which is known to increase portal flow. The second group (II) consisted of 31 patients with cirrhosis who received a dose of propranolol which is known to decrease portal flow. In Group I, 30 min after the meal, the portal vein blood velocity increased by 35 +/- 6% (p less than 0.01) in cirrhotic patients and by 55 +/- 5% (p less than 0.01), in normal subjects. The portal vein cross-sectional area increased significantly in normal subjects (22 +/- 2%, p less than 0.01) but not in cirrhotic patients (4 +/- 2%, n.s.). In Group II, 2 h after propranolol, there was a significant decrease in portal blood velocity (-14 +/- 2%), whereas the portal vein cross-sectional area did not show any significant changes. These data demonstrate that, in portal hypersensitive patients, the portal area measured by echo-Doppler flowmetry can be assumed to be constant and hence its calculation to estimate changes in portal blood flow can be omitted. Therefore, the use of blood velocity alone is suggested to monitor acute changes in flow in portal hypertension using Doppler flowmetry. The elimination of the portal vein cross-sectional area measurement simplifies the quantitative calculation of portal hemodynamics and increases the reliability of the technique by avoiding a source of error.