Patients undergoing pancreaticoduodenectomy (PD) often require postoperative artificial nutrition. This trial was undertaken to evaluate whether the route of administration and the composition of the postoperative nutritional support could affect the immunometabolic response and outcome. A prospective, randomized trial was carried out in 212 subjects who underwent PD. Patients were randomized to receive a standard enteral formula (standard group, n = 73) or an enteral formula enriched with arginine, omega-3 fatty acids, and RNA (immunonutrition group, n = 71), or total parenteral nutrition (parenteral group, n = 68). Postoperative feeding started 6 hours after surgery. The three regimens were isocaloric and isonitrogenous. Assessed parameters were phagocytosis ability of polymorphonuclear cells, plasma interleukin-2 receptors, C-reactive protein, retinol binding protein, tolerance of enteral feeding, rate of postoperative complications, and length of hospital stay (LOS). Full nutritional goal (25 kcal/kg) was achieved in 87% of enterally fed patients versus 95% in the parenteral group. Subjects receiving immunonutrition had a significantly better recovery of the immunometabolic parameters on postoperative day 8 compared to the other two groups. The rate of postoperative complications was lower in the immunonutrition group (33.8%) than in either the standard (43.8%) or parenteral group (58.8%) (p = 0.005 immunonutrition vs. parenteral). Also, the mean LOS was shorter in the immunonutrition group than in the standard and parenteral groups (15.1 vs. 17.0 vs. 18.8 days, respectively; p < 0.05). Early postoperative enteral feeding may safely and effectively replace parenteral nutrition in patients undergoing PD. Immunonutrition ameliorates the immunometabolic response and improves outcome compared to parenteral feeding.