A randomized trial was undertaken to determine the efficacy of nitroglycerin in addition to terlipressin infusion to improve the control of acute variceal hemorrhage compared with balloon tamponade. Forty-seven bleeding episodes in 34 cirrhotic patients were included, with terlipressin plus sublingual nitroglycerin in 23 episodes (group I) and balloon tamponade in 24 episodes (group II). At the end of the 12-hr period, hemorrhage had been equally controlled in both groups (18 of 23; 78% in group I and 19 of 24; 79% in group II). When attempted, balloon tamponade appeared more efficient in the failures of group I (4 of 4) than did terlipressin plus nitroglycerin in the failures of group II (0 of 3), although this difference was not significant. Major complications were rare (one in each group) and never required cessation of therapy. Thus terlipressin and nitroglycerin were as effective as balloon tamponade in controlling variceal hemorrhage and were associated with few minor complications.