Background This study aims to investigate the noninferiority of Enhanced Recovery After Surgery (ERAS) for pancreaticoduodenectomy (PD). Methods In this single‐center trial, we randomly assigned 276 adult patients who underwent open PD into ERAS and conventional groups with 138 patients in each, from 2015 through 2017. The primary endpoint was the incidence of overall morbidity until postoperative 3 months. The secondary endpoints were in‐hospital or 30‐day mortality, postoperative length of stay (LOS), nutritional status and overall hospital costs. Results Overall morbidity was reported in 64 patients (52.0%, ERAS group) and in 68 patients (54.8%, conventional group) (risk difference RD −2.81 percentage points (pp); 90% two‐sided confidence interval −13.24 to 7.63). Mortality did not occur in any patients. The two groups did not differ significantly in median postoperative LOS (both 11 days; RD −8.46 pp), body mass index (22.4 ± 2.75 vs. 22.4 ± 2.65 kg/m2; RD −3.48 pp), Patient‐Generated Subjective Global Assessment score over 4 (45 40.5% vs. 50 43.1% patients; RD −2.56 pp), and median overall hospital cost (15.61 vs. 16.04, ×106 KRW; RD −6.08 pp). Conclusions Even in PD, modified ERAS protocol was not inferior to conventional protocol, while reducing treatment burden. Highlight This randomized clinical trial by Hwang and colleagues revealed that the Enhanced Recovery After Surgery (ERAS) protocol was not inferior to the conventional perioperative management protocol for pancreaticoduodenectomy regarding morbidity, mortality, length of stay and hospital cost. The ERAS protocol not only reduces treatment burden, but also facilitates treatment administration.