OBJECTIVES: Preliminary studies suggested that octreotide may be therapeutic in bleeding angiodysplasia. Our aim was to investigate the efficacy of long-term octreotide therapy in the prevention of rebleeding from gastrointestinal angiodysplasia. METHODS: A cohort of 32 patients diagnosed with bleeding from angiodysplasia was treated with octreotide 50 μ 12 h subcutaneously for a 1-2 yr period. This cohort was compared with an external control group (38 patients who had received placebo [1 tablet/day] in a concurrent randomized clinical trial for the same period. RESULTS: Two patients of the octreotide group were lost to follow-up. Treatment failure occurred in seven of 30 (23%) patients in the octreotide group and in 17 of 35 (48%) in the placebo group (three dropouts before first visit) (P = 0.043). The actuarial probability of remaining free of rebleeding at 1 and 2 yr of follow-up was 77% and 68%, respectively, for the octreotide group and 55% and 36%, respectively, for the placebo group (log rank P = 0.030). Multivariate proportional hazards-regression analysis showed that octreotide therapy and previous bleeding episodes were positive and negative predictors of efficacy, respectively. No significant differences between the groups were observed according to number of bleeding episodes (0.4 ± 0.7 vs 0.9 ± 1.5, P = 0.070) and transfusion requirements (1.1 ± 2.6 vs 0.7 ± 1.5 units); however, iron requirements were lower in the octreotide than in the placebo group (22 ± 62 vs 166 ± 267 units; P < 0.001). Likewise, major adverse events (1 vs 1) and mortality (0 vs 1) were similar between groups. CONCLUSIONS: This study suggests that octreotide treatment may be beneficial in preventing rebleeding from gastrointestinal angiodysplasia. © 2007 by Am. Coll. of Gastroenterology.