The high mortality and wide variation in treatment of massive upper gastrointestinal hemorrhage indicate that its management is far from satisfactory. Therefore, Wangensteen's introduction of gastric hypothermia for control of acute hemorrhage has seemed a welcome innovation.9 Two other clinics have published experience with this procedure, citing nine7 and ten4 cases, respectively, and expressing encouragement although eight of these patients died. This new procedure has been employed only in the most seriously ill, so that a high mortality at first may be expected. The value of gastric hypothermia probably can be assessed only if its impact is measured on an entire institution's experience with serious gastrointestinal hemorrhage. These considerations, plus dissatisfaction with all present types of emergency operation for the control of hemorrhage, have led to a review of the entire experience with this problem at the University Hospital from 1950 through 1962. This interval provides experience