Objective To investigate the application value of Braun anastomosis in pancreatico-duodenectomy. Methods The retrospective cohort study was conducted. The clinicopathological data of 389 patients who underwent pancreaticoduodenectomy in the Fudan University Shanghai Cancer Center from March 2012 to July 2014 were collected. Of 389 patients, 235 receiving Braun anastomosis and 154 receiving non-Braun anastomosis were respectively allocated into Braun anastomosis group and non-Braun anastomosis group. All the patients underwent pancreaticoduodenectomy with digestive tract reconstruction using Child method. Patients in the Braun anastomosis group received 5-10 cm Braun anastomosis between input and output end of jejunum, and patients in the non-Braun anastomosis group didn′t receive jejunum-jejunum Braun anastomosis after gastrojejunostomy. Observation indicators included: (1) intraoperative situations; (2) postoperative recovery; (3) follow-up. Patients were followed up using outpatient examination and telephone interview up to May 2015. Follow-up included monthly routine blood retest, hepatorenal function retest and urine and stool routine retest, and enhanced CT scan in the epigastric region for every three months to detect recovery of digestive tract function. Measurement data with normal distribution were represented as ±s. Comparison between groups was analyzed using t test, and count data were analyzed using chi-square test. Results (1) Intraoperative situations: 389 patients underwent successful pancreaticoduodenectomy. Standard pancreaticoduodenectomy and pyloric-preserving pancreaticoduodenectomy were respectively applied to 205 and 30 patients in the Braun anastomosis group and 137 and 17 patients in the non-Braun anastomosis group, with no statistically significant difference (χ2=0.259, P>0.05). Anastomosis and reconstruction of pancreatic stump: anastomosis of main pancreatic duct and jejunal mucosa, embedded anastomosis of papillary main pancreatic duct and pancreas-stomach anastomosis were detected in 138, 89, 8 patients in the Braun anastomosis group and 85, 60, 9 patients in the non-Braun anastomosis group, respectively, with no statistically significant difference (χ2=1.535, P>0.05). Total operation time, pancreas-jejunum anastomosis time and volume of intraoperative blood loss were (398.9±61.9)minutes, (20.6±3.5)minutes, (401±59)mL in the Braun anastomosis group and (401.3±59.2)minutes, (20.7±2.1)minutes, (407±159)mL in the non-Braun anastomosis group, respectively, with no statistically significant difference (t=-0.380, -0.562, -0.319, P>0.05). (2) Postoperative recovery: time to initial anal exsufflation, time for fluid diet intake and time of drainage tube removal were (103±28)hours, (77±25)hours, (12±5)days in the Braun anastomosis group and (102±31)hours, (79±30)hours, (13±6)days in the non-Braun anastomosis group, respectively, with no statistically significant difference (t=0.330, -0.712, -1.783, P>0.05). Delayed gastric emptying, gastrointestinal hemorrhage, obstruction of afferent loop and pancreatic fistula were detected in 25, 3, 0, 30 patients in the Braun anastomosis group and 27, 4, 2, 23 patients in the non-Braun anastomosis group, respectively, with no statistically significant difference (χ2=3.818, 0.918, 3.068, 0.695, P>0.05). Seventeen patients were combined with delayed gastric emptying and pancreatic fistula, including 8 in the Braun anastomosis group and 9 in the non-Braun anastomosis group, with no statistically significant difference between the 2 groups (χ2=1.363, P>0.05). Patients with postoperative complications were improved by symptomatic and supporting treatment. Duration of hospital stay and treatment expenses were (14±7)days, (73 205±4 538)yuan in the Braun anastomosis group and (22±11)days, (83 219±5 738)yuan in the non-Braun anastomosis group, with statistically significant differences between the 2 groups (t=-8.767, -19.139, P<0.05). (3) Follow-up: 389 patients were followed up for 6 months, without death. Six and 9 patients in the Braun anastomosis group and non-Braun anastomosis group had regurgitation cholangitis. There was no readmission due to gastrointestinal hemorrhage and digestive tract obstruction, and no signs of hyperglycaemia and intractable diarrhea occurred. Conclusion Braun anastomosis can reduce duration of postoperative hospital stay and treatment expenses. Key words: Pancreaticoduodenectomy; Braun anastomosis; Delayed gastric emptying; Pancreatic fistula