A total of 63 patients with variceal bleeding were included in this study. Fifty-six attempts at percutaneous transhepatic variceal obliteration were made using stainless steel coils followed by 50% glucose and Gelfoam in 27 emergency cases, in whom bleeding did not stop by conventional medical treatment; and in 18 elective cases, in whom bleeding did stop by conventional medical treatment. The remaining 18 patients, whose bleeding was controlled by conventional medical treatment, were used as a control for the elective cases (conservative cases). The overall success rate was 93%. In 92% of the 37 acute bleeders, bleeding ceased as soon as the varices were obliterated. In emergency cases, the cumulative variceal rebleeding rate at 1, 2, 3, 6, 9, and 12 months after obliteration was 16%, 29%, 34%, 44%, 56%, and 56%, respectively. Its mortality within 1 month after the first bleeding was only 11%. In elective cases, the rebleeding rate at 1, 2, and 12 months was significantly lower; and the survival rate at 1 and 2 months was significantly higher compared with conservative cases. Follow-up portography in 10 active rebleeders and two nonrebleeders demonstrated new vessel formation in six, and recanalization of previously completely occluded varices in two. Complications included transient hemiparesis and partial stenosis of intrahepatic portal branches, but none was fatal. When compared with a conventional treatment, transhepatic variceal obliteration using steel coils followed by 50% glucose and Gelfoam proved to be an effective, safe emergency treatment for variceal hemorrhage. However, since the rebleeding rate was high, this procedure should be followed by an elective operation or other procedures for a lasting prevention of bleeding.