This study aims to assess the long-term effects of bariatric surgery on type 2 diabetic patients. We searched Cochrane Library, PubMed, and EMbase up to Dec 2013. Randomized controlled trials (RCTs) and cohort studies of bariatric surgery for diabetes patients that reported data with more than 2 years of follow-up were included. We used rigorous methods to screen studies for eligibility and collected data using standardized forms. Where applicable, we pooled data by meta-analyses. Twenty-six studies, including 2 RCTs and 24 cohort studies that enrolled 7883 patients, proved eligible. Despite the differences in the design, those studies consistently showed that bariatric surgery offered better treatment outcomes than non-surgical options. Pooling of cohort studies showed that BMI decreased by 13.4 kg/m(2) (95 % confidence interval (CI), -17.7 to -9.1), fasting blood glucose by 59.7 mg/dl (95 % CI, -74.6 to -44.9), and glycated hemoglobin by 1.8 % (95 % CI, -2.4 to -1.3). Diabetes was improved or in remission in 89.2 % of patients, and 64.7 % of patients was in remission. Weight loss and diabetes remission were greatest in patients undergoing biliopancreatic diversion/duodenal switch, followed by gastric bypass, sleeve gastrectomy, and adjustable gastric banding. Bariatric surgery may achieve sustained weight loss, glucose control, and diabetes remission. Large randomized trials with long-term follow-up are warranted to demonstrate the effect on outcomes important to patients (e.g., cardiovascular events).