BACKGROUND: Several studies have found that metabolic surgery can significantly improve glucose homeostasis; however, the intrinsic mechanisms remain unclear. Accumulating evidence suggests that duodenal bypass plays a crucial role in the treatment of type 2 diabetes mellitus (T2DM). Here, we aimed to evaluate the effect of duodenal reflux on glucose metabolism in T2DM. METHODS: A high-fat diet and low-dose streptozotocin (STZ) administration were used to induce T2DM in male rats, which were assigned to three experimental groups: sham operation (SO; n = 10), new duodenal-jejunal bypass (NDJB; n = 10), and new duodenal-jejunal bypass with a tube (NDJBT; n = 10). Weight, food intake, oral glucose tolerance test (OGTT) results, glucagon-like peptide 1 (GLP-1) levels, and histopathology were assessed before or after surgery. Plain abdominal radiography was performed 1 week after the operation. RESULTS: Plain abdominal radiography indicated the occurrence of contrast agent reflux into the duodenum. The body weight and food intake in all three groups did not significantly differ before and after surgery. The NDJB and particularly the NDJBT groups exhibited better glucose tolerance, lower fasting blood glucose (FBG) levels, lower area under the curves for OGTT (AUC) values, and higher GLP-1 levels, as compared with the sham group postoperatively. The villus height and crypt depth were both shorter in the biliopancreatic limb after NDJBT, as compared with those after SO and NDJB. CONCLUSIONS: Thus, exclusion of the duodenum alone and tube placement can effectively prevent duodenal reflux and improve glucose homeostasis, which further suggests that the duodenum plays an important role in T2DM.