Background This study was designed to compare surgical, morphological, and functional outcomes of pancreaticoduodenectomy (PD) according to the types of pancreaticoenterostomy performed and to suggest a proper anastomotic method after PD. Methods From January 2001 to December 2006, 147 PDs were performed at Ajou University Medical Center. Surgical, morphological, functional, and nutritional outcomes after PD were retrospectively compared according to the types of management of pancreatic remnant and whether pancreaticogastrostomy (PG) or pancreaticojejunostomy (PJ), including duct-to-mucosa or invagination method, was performed. Results For the reconstruction method, 43 PG (30 ductto-mucosa and 13 invagination) and 100 PJ (33 duct-tomucosa and 67 invagination) were performed. Pancreatic leak rate in PG group (7%) was less than that in PJ group (13%); however, it was not significant (P [ 0.05). On the other hand, there was a significant difference in pancreatic leak between duct-to-mucosa and invagination (3.2 vs. 17.5%, P \ 0.05). Surprisingly, there was no pancreatic leak in PG duct-to-mucosa anastomosis after PD. There were no significant differences in the change of remnant pancreatic duct size, pancreatic thickness, presence of steatorrhea, and new-onset diabetes mellitus (DM) between PG and PJ. In the invagination group, the main pancreatic duct diameter was increased and pancreatic thickness was progressively reduced. Conclusion The duct-to-mucosa method is safer and has a good duct patency and low pancreas atrophy compared with the invagination method. In addition, PG duct-tomucosa is safer than PG invagination, but not in the PJ group. Therefore, we recommend PG duct-to-mucosa for reconstruction after PD because of safety and good duct patency, especially for inexperienced surgeons.