The efficacy of antifibrinolytic therapy in the management of acute upper gastrointestinal haemorrhage has been investigated in a double-blind clinical trial. Two-hundred patients were studied using tranexamic acid, a potent antifibrinolytic agent. Of these, 103 were in the treatment group and 97 in the control group. Patients were analysed to determine severity of initial blood loss, transfusion requirements, together with the incidence of recurrent bleeding, surgical intervention, and death. Final diagnosis as to the site of bleeding was arrived at using endoscopy, barium studies, and the findings at operation and necropsy. The groups were well matched as regards severity of initial haemorrhage, age, sex, aetiological diagnosis, and precipitating factors. A significant difference was observed in the requirement for surgical intervention to control continuing or recurrent haemorrhage. Twenty-three of 97 in the control group and seven of 103 in the treatment group required surgery. There appeared to be a reduction in the transfusion rate after the first three days of hospitalization in the treatment group. There were no significant differences in mortality or in side-effects between the two groups.