A multicenter double-blind clinical trial was undertaken to evaluate the efficacy of a short-term somatostatin treatment versus a short-term vasopressin treatment on acute hemorrhage from esophageal varices in patients with liver cirrhosis and portal hypertension. Forty-nine patients with massive hemorrhage and endoscopic diagnosis of bleeding esophageal varices completed the study. Patients were randomly assigned to somatostatin treatment (24 patients: 250 micrograms/hr i.v. for 48 hrs) or vasopressin treatment (25 patients: 0.1 U/min i.v. for 48 hrs). The Sengstaken-Blakemore tube was utilized, when needed, for a six hour period. In case of failure the patients were crossed-over to the other treatment. Patients in whom the bleeding stopped at 48 hrs, were randomly assigned to somatostatin (250 micrograms/hr i.v.) or placebo for seven days. Bleeding stopped in 68% of patients treated with somatostatin and in 28% of patients treated with vasopressin (p less than 0.0013). Mortality rate was lower, but not significantly so, in the somatostatin group compared to the vasopressin group. No differences were noted between somatostatin and placebo in preventing bleeding recurrences. These data suggest that somatostatin, when combined if necessary with a 6 hour period of balloon tamponade, is more effective than vasopressin at low doses in controlling severe hemorrhage from esophageal varices in patients with liver cirrhosis and portal hypertension. A clinical use of somatostatin seems to be indicated in these patients.