Perioperative fluid restriction has been suggested to reduce morbidity and length of stay. The purpose of this study was to compare the morbidity following pancreaticoduodenectomy (PD) between fluid restriction group and conventional management group. Seventy-two patients were enrolled for perioperative fluid restriction of PD. During the operation, main fluid was infused at a rate of less than 8mL/kg/hr. Until POD#3, 10% dextrose and Hartmann's solution were administered at rates of 40 mL/h and {(1.5*body weight) - 42} mL/h, respectively. The historical control group consisted of 139 patients. We compared the rates of major complication (Clavien-Dindo grade III to V) and clinically relevant postoperative pancreatic fistula (CR-POPF), length of hospital stays (LOS), amount of urine output, and the rate of acute kidney injury (AKI). The rates of major complication (19.0% versus 18.7%; p > 0.999), CR-POPF (15.5% versus 15.1%; p > 0.999), and LOS (19 days [range: 10-52] versus 19 days [range: 11-75]; p=0.514) were comparable in the study and the control group, respectively. Amount of urine output during the operation and from POD#1 to POD#3 was more than minimal amount (0.5mL/kg/hr) in the both groups. Incidence rate of AKI in the study group was not higher than the control group (Stage I: 1.7% versus 2.9%, p > 0.999; stage II: 0% versus 1.4%, p > 0.999). There was no decrease in incidence of morbidity including POPF following PD with perioperative fluid restriction. Fluid restriction was feasible because it did not reduce urine output and did not increase incidence of AKI.