We conducted a prospectively randomized trial of 78 patients with peptic ulcer hemorrhage to evaluate the hemostatic effects of heat probe thermocoagulation and pure alcohol injection. The initial and ultimate success rates of heat probe thermocoagulation were better than those of pure alcohol injection (p less than 0.05). Rebleeding rates and success rates of retreatment in the two groups were not significantly different. In the case of a spurter or a bleeder located over the lesser curvature side of the stomach or superior wall of the duodenal bulb, heat probe thermocoagulation was better than pure alcohol injection in achieving hemostasis. We conclude that heat probe thermocoagulation is better than pure alcohol injection in arresting peptic ulcer hemorrhage. If the bleeder is a spurter or is located over the lesser curvature side of the stomach or superior wall of the duodenal bulb, heat probe thermocoagulation is the treatment of choice.