The aim of this study is to ascertain role of respiratory-gated PET/CT for accurate diagnosis of pancreatic tumors. Prior to clinical study, the phantom study was performed to evaluate the impact of respiratory motion on lesion quantification. Twenty-two patients (mean age 65 years) with pancreatic tumors were enrolled. Pathological diagnoses by surgical specimens consisted of pancreatic cancer (n=15) and benign intraductal papillary mucinous neoplasm (IPMN, n=7). Whole-body scan of non-respiratory-gated PET/CT was performed at first, and subsequent respiratory-gated PET/CT for one bed position was performed. All PET/CT studies were performed prior to surgery. The SUV max obtained by non-respiratory-gated PET/CT and respiratory-gated PET/CT, and percent difference in SUVmax (%SUVmax) were compared. The profile curve of 5 respiratory bin image was most similar to that of static image. The third bin of 5 respiratory bin image showed highest FWHM (24.0mm) and FWTM (32.7 mm). The mean SUVmax of pancreatic cancer was similar to that of benign IPMN on non-respiratory-gated PET/CT (p=0.05), whereas significant difference was found between two groups on respiratory-gated PET/CT (p=0.016). The mean %SUV of pancreatic cancer was greater than that of benign IPMN (p<0.0001). Identification of the primary tumor in pancreatic head (n=13, 59%) was improved by using respiratory-gated PET/CT because of minimal affection of physiological accumulation in duodenum. Respiratory-gated PET/CT is a feasible technique for evaluation of pancreatic tumors and allows more accurate identification of pancreatic tumors compared with non-respiratory-gated PET/CT. Copyright © 2012 Elsevier Ireland Ltd. All rights reserved.