Purpose To determine useful magnetic resonance (MR) imaging features to differentiate nonhypervascular pancreatic neuroendocrine tumors (PNETs) from pancreatic ductal adenocarcinomas (PDACs). Materials and Methods The institutional review board approved this retrospective study and waived the informed consent requirement. Seventy-four patients with surgically confirmed PNETs and 82 patients with PDACs who underwent gadobutrol-enhanced MR imaging were included. Two radiologists independently evaluated the morphologic characteristics and temporal enhancement patterns of each tumor. Quantitative analysis, including measurement of tumor size, maximal upstream parenchymal thickness (MUPT), contrast-to-noise ratio, and apparent diffusion coefficient values, was performed. Uni- and multivariate logistic regression analyses were performed to identify relevant features to differentiate between PNETs and PDACs. Results On the basis of arterial enhancement, 38 PNETs (51%, 38 of 74) were hypervascular and 36 PNETs (49%, 36 of 74) were nonhypervascular. At MR imaging, nonhypervascular PNETs showed significantly higher frequencies of a well-defined margin, portal hyper- or isoenhancement, and MUPT of 10 mm or greater but lower frequencies of ductal dilatation, vascular invasion, and peripancreatic infiltration when compared with PDACs (P < .05 for all). At multivariate analysis, a well-defined margin and portal hyper- or isoenhancement were independent significant differentiators of PNETs from PDACs (odds ratio, 20.3 and 16.1, respectively). When applying the criteria of a well-defined margin and portal hyper- or isoenhancement, 64% of sensitivity and 99% of specificity were observed for the differential diagnosis of PNETs from PDACs. Conclusion A well-defined margin and hyper- or isoenhancement in the portal venous phase are useful MR imaging features that are more common in nonhypervascular PNETs and may help discriminate nonhypervascular PNETs from PDACs. © RSNA, 2017 Online supplemental material is available for this article.