Background: This study assessed the long-term effectiveness of push enteroscopic cauterization of bleeding intestinal angiodysplasia. Methods: We retrospectively reviewed the clinical course of patients who underwent push and sonde enteroscopy for obscure gastrointestinal bleeding and were diagnosed with intestinal angiodysplasias. Results: One hundred twelve patients bleeding from small intestinal angiodysplasias were identified. After excluding those lost to follow-up (29), data were collected from 83 patients. Fifty-five patients (29 men; mean age, 73 years; mean units of packed red blood cells transfused, 21.4; average bleeding history, 22 months) were cauterized. Twenty-eight patients (12 men; mean age, 71; mean units of packed red blood cells transfused, 15.8; average bleeding history, 22 months) were not cauterized. The noncauterized group (follow-up, 26 ± 14 months; mean ± SD) continued to bleed, requiring 2.16 ± 3.86 units of packed red blood cells transfused per month (units/month) before and 0.97 ± 1.46 units/month after diagnosis (NS). The cauterized group (follow-up, 30 ± 18 months) significantly improved, requiring 2.40 ± 2.97 units/month before treatment and 0.32 ± 0.91 units/month after cauterization (p < 0.0001, paired t test). Conclusion: Cauterization of endoscopically accessible small intestinal angiodysplasias may decrease rebleeding. (Gastrointest Endosc 1996;43:580-3.)