Diabetes mellitus (DM) is a public health problem with a prevalence of 345 million people worldwide that it may double by the year 2030 and have a high costs and mortality. Gastrointestinal surgery is accepted as a form of treatment that was already suggested for obese in 1987 by Pories, confirmed for obese patients by the meta-analysis of Buchwald and the direct comparison of gastric bypass with medical treatment in the study of Schauer that demonstrate a 4 fold greater resolution rate of DM with surgery. Improvement occurs immediately after surgery, before the patients lose weight in with BMI > 35; but there is doubt if the existent evidence is enough to extrapolate these results to patients with BMI < 35 and especially with BMI < 30, in spite that four reviews in patients with this BMI and DM2 demonstrated the same results when stomach, duodenum and part of jejunum is bypassed as happen gastric bypass (better results with this of one anastomosis than of two anastomosis, Rouxen-Y) BPD. For patients with a BMI between 30 and 35 restrictive techniques: LAGB and SGL are good but not better than the mixed: RYGB, BAGUA, or SG-DJB with remission from 60 to 100%, minor in the derivative: BPD and above on the IID with a 81% of remission. There are no differences in the metabolic control in comparison to the obese, It is progressively better with DJB, SDS, IID and BAGUA especially in patients who do not require insulin, have less time with disease, have normal C peptide levels, and not so much relation with the initial BMI that is only important to decide the degree of restriction. Although several mechanisms has been suggesed for explaining these results such as caloric intake, hormonal changes, bypass of the anterior or early stimulation of posterior intestine, fundectomy, intestinal gluconeogenesis and others, new ones will appear in the near future.