Background/aims: The mechanism of impaired glucose metabolism that develops in most patients with pancreatic cancer is obscure. The association between pancreatic cancer and diabetes is controversial. Impaired glucose tolerance or diabetes mellitus may develop as a clinical manifestation of pancreatic cancer; however, diabetes may be a predisposing risk factor for pancreatic cancer. We aimed to investigate the relationship between diabetes and pancreatic cancer, and also the impact of tumor removal on glucose metabolism. Methods: Eighteen pancreatic cancer patients with resectable tumors and without previous diabetes history were enrolled. All patients underwent oral glucose tolerance test and measurement of insulin levels before and after Whipple procedure. Results: Eight of 18 (44.4%) patients were diabetic before surgery whereas 4 (22.2%) had impaired glucose tolerance. Only 6 (33.3%) patients had normal glucose metabolism at the first clinical admission. After pancreatectomy, only 4 (22.2%) patients were diabetic and 1 (5%) had impaired glucose tolerance. Thirteen patients (72%) had normal glucose metabolism after tumor removal. In 8 patients, impaired glucose metabolism improved after surgery. Only 1 patient out of 6 (16%) with normal glucose metabolism initially developed impaired glucose tolerance after surgery. All patients with diabetes and impaired glucose tolerance had hyperinsulinemia before and after surgery. Insulin levels were lower after surgery than before surgery, and glucose metabolism was improved postoperatively. Conclusions: Our results showed that tumor removal in pancreatic cancer patients improved glucose metabolism. This occurred despite a postoperative reduction in endocrine pancreas mass, which may suggest the presence of insulin resistance and diabetogenic effect of pancreatic cancer. The elucidation of the mechanism is of immense importance for providing an early tumor marker and preventative and therapeutic modalities.