A double-blind, multicenter trial was carried out to assess the effectiveness of isosorbide-5-mononitrate in preventing recurrent variceal hemorrhage during the course of endoscopic sclerotherapy. Seventy-six patients with their first bleeding episode from esophageal varices were randomly allocated, after initial control of hemorrhage, to groups receiving either 50 mg/day oral isosorbide-5-mononitrate retard (37 patients) or an identical placebo (39 patients) until variceal eradication. Sclerotherapy was performed at weekly intervals, and varices were intra- and para-variceally injected with 1% polidocanol until eradication. If rebleeding occurred, additional sclerotherapy was performed. Four (10.8%) patients rebled in the isosorbide group, compared with 15 (38.4%) in the placebo group (p = 0.01). The total number of rebleeding episodes was also significantly lower in the isosorbide group (5 versus 19, p = 0.043), whereas comparison between major versus minor rebleedings was not significant. The median transfusion requirement per bleeding episode was not significantly different in the two groups, although the cumulative number of blood units transfused was over threefold greater (22 versus 70) in the placebo group. Two (5.4%) deaths occurred among isosorbide-treated patients and nine (17.9%) among placebo patients (NS). The number of sclerotherapy sessions and the time required to obtain variceal eradication were also comparable in the two groups. Finally, the nitrovasodilator was well tolerated, requiring withdrawal for severe headache in only one patient. In conclusion, isosorbide-5-mononitrate reduces the rebleeding rate and the number of rebleeding episodes before variceal eradication in patients treated with sclerotherapy.