Six hundred sixty-four patients who have had metabolic intestinal surgery for the treatment of morbid obesity are reviewed. Particular attention is directed to difference in weight loss, morbidity and mortality in end-to-side and end-to-end shunts, the former performed in 300 patients and the latter in 262 patients. Unsatisfactory weight loss was observed in 20% of patients with end-to-side shunts, while only 8% of patients with end-to-end shunts failed to lose a sufficient amount of weight. Morbidity and mortality were significantly increased in this latter group. This led to adoption of a modification of the end-to-side operation by developing a plication to prevent reflux. Our preliminary observation indicates that in 102 patients who have had this operation, similar weight loss is attained to that of end-to-end shunts with no greater morbidity or mortality than the end-to-side type.