Objective: To investigate the effects of two different decompression tube placement methods on pancreatic fistula after pancreaticoduodenectomy (PD). Methods: The clinical data of 125 patients undergoing PD from January 2012 to October 2014 were retrospectively analyzed. Of the patients, 68 cases underwent decompression by insertion of a drainage tube into the main pancreatic duct (pancreatic duct group), and 57 cases underwent decompression by insertion of a drainage tube into the distal part of the blind end of the pancreatic jejunal anastomosis (blind loop group). The incidence of postoperative pancreatic fistula and other clinical variables between the two groups were compared. Results: Pancreatic fistula occurred in 9 cases (7.2%) in the entire group, with one case (1.8%) in blind loop group and 8 cases (11.8%) in pancreatic group, and all cases were cured by conservative treatment and discharged from hospital. The incidence of pancreatic fistula in blind loop group was significantly lower than that in pancreatic duct group (P<0.05), but there was no statistical difference in incidence of other complications between the two groups (all P>0.05). In blind loop group compared with pancreatic duct group, the length of hospital stay (20.38 d vs. 27.46 d) and hospitalization cost (56 420 yuan vs. 66 020 yuan) were significantly reduced (both P<0.05). No statistical difference was noted in postoperative 1-year tumor recurrence and survival rate between the two groups (both P>0.05). Conclusion: The decompression tube placement via blind loop can effectively reduce the incidence of postoperative fistula after PD.