Molsidomine, a long acting vasodilator with antianginal properties, has been shown to decrease porto-hepatic pressure gradient in patients with cirrhosis. The present study aimed at assessing the effects of molsidomine, propranolol and of the association of these two drugs on portal vein blood flow as measured using Doppler and B-mode sonography. In 10 patients without liver disease (group 1), portal flow time average mean velocity (TAV) and portal vein blood flow (PVBF) were measured under basal conditions, 1 hour then 2 hours after ingestion of 4 mg of molsidomine. The same measurements were performed in 15 patients with cirrhosis (group 2) under basal conditions, 1 then 2 hours after double-blind administration of either molsidomine (10 patients) or placebo (5 patients). Fifteen further patients with cirrhosis (group 3) were studied after the double blind administration of 80 mg of propranolol and two hours later of 4 mg of molsidomine (10 patients) or placebo (5 patients); TAV and PVBF were measured under basal conditions, two hours after propranolol ingestion or placebo, then one and two hours after molsidomine or placebo ingestion. TAV and PVBF remained unchanged in patients treated with placebo. Molsidomine reduced TAV by 23.8 +/- 19.5% in group 1 (P < 0.01) and by 25.6 +/- 21.4% in group 2 (P < 0.01). In group 3, a 10% decrease was observed after propranolol (NS). When molsidomine was added, TAV was further decreased (-17.6 +/- 13.3% vs baseline, P < 0.01). PVBF remained unchanged in the three groups of patients.(ABSTRACT TRUNCATED AT 250 WORDS)