The aim of this study was to test the effectiveness of isosorbide-5-mononitrate (IM) as an adjunct to propranolol (PR) in the prevention of variceal rebleeding. Ninety-five cirrhotic patients with variceal bleeding were randomly assigned to treatment with PR + IM (46 patients) or PR alone (49 patients). Eighteen patients in the PR + IM group and 28 in the PR group had rebleeding during the 2 years after randomization. The actuarial probability of rebleeding 2 years after randomization was lower in the PR + IM group (40.4% vs. 57.4%) but the difference was not significant (P =. 09). However, the decrease in the risk of rebleeding reached statistical significance after stratification according to age, i.e. less than 50 versus >/=50 years old, (P =.03) or by adding an additional year of follow-up (P =.05). No significant difference was found in rebleeding index and survival. The multivariate Cox analysis indicated first, that both treatment (P =.03) and age (P =. 001) were factors predictive of rebleeding and second, that PR + MI reduced the risk of rebleeding by half (relative risk: 0.51, 95% confidence interval: 0.28-0.95). Seven patients in the PR + MI group and 1 patient in the PR group had to discontinue one of the drugs because of adverse events (P =.03). These results suggest that the addition of IM improves the efficacy of PR alone in the prevention of variceal rebleeding in cirrhotic patients. However no beneficial effects were observed on other parameters reflecting the efficacy of treatment.