Sixteen patients (15 males, aged 48-70) affected by liver cirrhosis and oesophageal varices were subjected to duplex-Doppler ultrasonographic study (DDUS). Four patients (three with a portal thrombosis and one with a hepatofugal portal flow) were excluded from the subsequent pharmacological test. The twelve remaining patients took part in a double blind cross-over study that evaluated the variations of heart rate (HR), mean systemic arterial pressure (SAP), portal vein diameter (PVD), maximal and mean portal flow velocity (PFV) after the administration of either 40 mg of propranolol or placebo per os, on two consecutive days. Propranolol caused no significant variation in mean SAP and in PVD, whereas it reduced the HR from 67.7 +/- 8.0 to 58.4 +/- 7.0 beats/min (mean +/- s.d.; P less than 0.001); the maxPFV dropped from 18.2 +/- 5.4 to 14.0 +/- 3.7 cm/s (P less than 0.001) and the meanPFV dropped from 15.3 +/- 4.1 to 13.2 +/- 3.1 cm/s (P less than 0.005). No significant variation was observed with placebo. After propranolol administration eight patients exhibited a significant maxPFV decrease, whereas the other four patients exhibited only a drop in HR, suggesting either drug inefficacy, inappropriate dosage or inadequate duration of treatment. DDUS is the only non-invasive method for the examination of the portal vein system.(ABSTRACT TRUNCATED AT 250 WORDS)