The present study was designed to investigate the effect of propranolol on portal pressure of patients with alcoholic cirrhosis and portal hypertension and to correlate these effects with clinical and laboratory parameters. The mean baseline hepatic venous pressure gradient in the 50 patients studied was of 18.2 +/- 4.1 mm Hg. It decreased significantly 2 hr after the oral administration of 40 mg of propranolol to 15.7 +/- 4.2 mm Hg (a mean reduction of 13.4 +/- 17%). This reduction in hepatic venous pressure gradient resulted mainly from a decrease in mean wedged hepatic venous pressure. There was no correlation between the decrease in hepatic venous pressure gradient and the decrease in heart rate. When results were analyzed individually, only 15 (30%) showed a large decrease in hepatic venous pressure gradient (greater than 20%), 15 (30%) showed a moderate decrease (10 to 19%), and in 20 patients (40%) there was no reduction or an increase in hepatic venous pressure gradient. Comparison of "responders" (those that reduced hepatic venous pressure gradient greater than 10%) and "nonresponders" (hepatic venous pressure gradient reduction less than 10%) showed no significant differences in baseline laboratory and hemodynamic parameters, in the severity of the liver disease, in the heart rate and blood pressure response to propranolol, nor in the propranolol plasma levels achieved 2 hr after propranolol administration. Propranolol plasma levels correlated with the reduction in heart rate but not with the reduction in hepatic venous pressure gradient. Of 14 nonresponders to 40 mg of propranolol who received additional doses, six showed a reduction in hepatic venous pressure gradient.(ABSTRACT TRUNCATED AT 250 WORDS)