Control of arterial bleeding from peptic uclers is a challenging clinical problem. Our purposes were to compare 1) effectiveness and safety of combination epinephrine injection (EPI) and bipolar coagulation (COMBO) vs. bipolar coagulation alone for hemostasis of actively bleeding peptic ulcers and 2) effectiveness of a new inject-Gold probe catheter. Methods: This is a multicenter, randomized, blinded study of high risk patients with clinically severe hemorrhage and active, arterial type bleeding from gastric or duodenal ulcers at emergency endoscopy. Patients with oozing bleeding were excluded. Patients were randomly assigned to endoscopic treatment with COMBO (injection in 4 quadrants of 1:10,000 EPI and Gold probe coagulation-10F, 15-20W, firm tamponade & 10 sec pulses) or bipolar alone (Gold probe-GP-without EPI). If bleeding could not be controlled within 15 min of endoscopic therapy, patients were classified as a failure and crossed over or sent to surgery. Patients were managed by a separate team of physicians who were blinded to the endoscopic treatment type. All patients received PPI's BID. Results: 36 pts (19 GP & 17 COMBO) were randomized. The two groups were similar at study entry except that signficantly more COMBO pts were ingesting NSAIDs (41% vs 11%) and the mean baseline Hct's were higher (26 vs. 22) than the GP group. See table for results. Primary hemostasis rate was significantly higher with COMBO. There were no deaths or complications. Inject-GP was rated by all investigators as faster and more convenient than a separate injection needle and GP. Conclusions. For bleeding ulcers with active arterial hemorrhage: 1) There were significantly higher rates of initial hemostasis with combination injection and bipolar coagulation than bipolar alone. 2) The new inject- Gold probe was rated by endoscopists as faster and more convenient than separate catheters. Funded by NIH DK33273 and partially by Microvasive-Boston Scientific.