Background/Purpose: Delayed gastric emptying (DGE) is one of the most common complications after pancreaticoduodenectomy (PD). The aim of the present metaanalysis was to evaluate the effect of Billroth-II(B-II) versus Roux-en-Y (R-Y) reconstruction for gastrojejunostomy on DGE after PD. Methods: A systematic literature search was performed using the electronic database MEDLINE (via PubMed and OVID), EMBASE and Cochrane Central Register of Controlled Trials (CENTRAL) of the Cochrane Library to select pertinent randomized controlled trials (RCTs) on this topic from January 1990 to January 2020.The primary outcome was identified as postoperative DGE. Subgroup analysis was established to compare the incidence of grade B and C DGE. Software Revman 5.3 was used for the statistical analysis, summary statistics were calculated using fixed effect model or random effect model. Results: Five RCTs including a total of 612 patients were eligible for this metaanalysis. The incidence of grade B and C DGE was significantly lower with the B-II reconstruction than with the R-Y reconstruction (8.0% vs. 14.8%, OR = 0.49, 95% CI: 0.26-0.95, P = 0.03) and the B-II reconstruction took a shorter operation time (WMD=−7.18, 95% CI: [−13.09, −1.27], P = 0,02). No statistically significant difference was found between the two reconstruction methods in terms of the incidence of postoperative pancreatic fistula (POPF), bile leak, intra-abdominal abscess, postoperative pneumonia and the length of postoperative hospital stay. Conclusions: B-II reconstruction after PD has a lower incidence of grade B and C DGE and shorter operation time compared with R-Y reconstruction.