The haemostatic effect of terlipressin (triglycyl-lysine vasopressin; Glypressin) on bleeding from oesophageal varices was evaluated in a placebo-controlled, double-blind, randomized clinical trial. Patients with clinically suspected liver cirrhosis were included in the study if they had been admitted to hospital with an extensive haemorrhage within the last 24h before diagnostic endoscopy. The patients randomized after stratification for severity of liver disease. Terlipressin or placebo was administered as intravenous bolus injections every 4th h during a period of 24 to 36 h or until the clinical course necessitated active intervention (failure or withdrawal). Sixty patients entered the study; 31 patients were allocated to receive terlipressin, and 29 patients to receive placebo. Bleeding from varices was arrested in 28 of the 31 receiving terlipressin, as compared with 17 of the 29 receiving placebo (p less than 0.01). Patients receiving active drug required significantly fewer blood transfusions (p less than 0.05). Most of the side effects were classified as mild and were registered in the terlipressin group.