Since February 1, 1980, 515 morbidly obese patients have undergone the Greenville gastric bypass (GGB) operation. Of these, 212 (41.2%) were euglycemic, 288 (55.9%) were either diabetic or had glucose intolerance, and 15 (2.9%) were unable to complete the evaluation. After the operation, only 30 (5.8%) patients remained diabetic (and 20 of these improved), 457 (88.7%) became and have remained euglycemic, and inadequate data prevented classification of the other 28 (5.4%). The patients who failed to return to normal glucose values were older and their diabetes was of longer duration than those who did. The effect of the GGB was not only limited to the correction of abnormal glucose levels. The GGB also corrected the abnormal levels of fasting insulin and glycosylated hemoglobin in a cohort of 52 consecutive severely obese patients with non-insulin-dependent diabetes. The GGB effectively controls weight. If morbid obesity is defined as 100 pounds over ideal body weight, 89% of the patients are no longer "morbidly" obese within 2 years. In most patients, the control of the weight has been well maintained during the 11 years of follow-up; most of the upward creep in weight of 20.8% between 24 and 132 months was from the 49 (9.5%) patients who had staple line breakdowns between the large and small gastric pouches. Non-insulin-dependent diabetes, previously considered a chronic unrelenting disease, can be controlled in the severely obese by the gastric bypass. Whether the correction of glucose metabolism affects the complications of diabetes is unknown. Whether the gastric bypass should be considered for patients with advanced non-insulin-dependent diabetes but who are not severely obese deserves consideration. The GGB has an unacceptably high rate of staple line failure. Accordingly, the authors have recently changed their procedure to one that divides the stomach rather than partitions it with staples.