BACKGROUND Delayed gastric emptying (DGE) is one of the main complications after pancreaticoduodenectomy (PD). Literature review and meta-analysis were used to evaluate whether subtotal stomach-preserving pancreaticoduodenectomy (SSPPD) may have less incidence than pylorus-preserving pancreaticoduodenectomy (PPPD). METHODS Online search for studies comparing PPPD to SSPPD was done. Primary outcome was DGE. Quality of included studies was evaluated and heterogeneity was assessed. Relative risk (RR) and 95% confidence intervals (CI) were calculated from pooled data in RCTs and retrospective studies. RESULTS Eight studies met our selection criteria, with a total of 663 patients undergoing pancreaticoduodenectomy; 309 underwent PPPD and 354 underwent SSPPD. Median age was 66 years. Average male/female ratio was 57 vs. 43%, respectively. There was lower incidence of DGE with SSPPD (RR 0.527; 95% CI 0.363-0.763; p < 0.001) and less nasogastric tube days with SSPPD (RR -0.544; 95% CI -876 to -0.008; p = 0.047). Operative blood loss was more in SSPPD (RR 0.285; 95% CI 0.071-0.499; p = 0.009). There was no statistical difference between the two groups regarding length of hospital stay, incidence of pancreatic fistula, abscesses, overall morbidity, or postoperative mortality. CONCLUSIONS SSPPD was associated with less DGE than PPPD. Larger prospective randomized studies are needed to investigate the association of this result with other complications in more depth.