We assessed the value of selective arteriography in the diagnosis and management of acute gastrointestinal hemorrhage.We reviewed the records of 107 consecutive patients who had gastrointestinal hemorrhage and underwent selective arteriography between January 1992 and October 2003: 10 had upper gastrointestinal bleeding, 79 had lower gastrointestinal bleeding, and 18 had varicose bleeding with portal hypertension. Selective embolization was attempted in 15 patients to obtain hemostasis. Angiographic findings were reviewed and prospective reports were compared with the final diagnosis and outcome.Of 129 angiographic studies, 36 correctly revealed the bleeding site and 93 were negative. Extravasation was seen in 24 cases at the level of stomach (n = 2), duodenum (n = 1), small bowel (n = 5), or colon (n = 16). Indirect signs of bleeding sources were identified in 12 patients (stomach in one, small bowel in four, large bowel in four, liver in three). Transcatheter embolization induced definitive hemostasis in 11 of 15 patients (73%), namely in the stomach (n = 2), small bowel (n = 3), colon (n = 7), and liver (n = 3). Three patients required surgery after embolization.Abdominal arteriography may localize gastrointestinal bleeding sources in approximately one-third of cases. Selective embolization may provide definitive hemostasis in most instances.