STUDY OBJECTIVE: To determine if the normobaric oxygen paradox (NOP) was effective in increasing reticulocyte count and reducing postoperative requirements for allogeneic red blood cell transfusion after traumatic hip surgery. DESIGN: Prospective, randomized, double blinded, multi-center study. SETTING: Surgical wards of two academic hospitals. PATIENTS: 85 ASA physical status 1 and 2 patients undergoing surgery for traumatic hip fracture. INTERVENTIONS: Patients were randomly assigned to receive 30 minutes of air [air group (control); n = 40] or 30 minutes of 100% oxygen (O(2) group; n = 14) at 15 L/min every day from the first postoperative day (POD 1) until discharge. MEASUREMENTS: Venous blood samples were taken at admission and after surgery on POD 1, POD 3, and POD 7. Hemoglobin (Hb), hematocrit (Hct), reticulocytes, hemodynamic variables, and transfusion requirements were recorded, as were hospital length of stay (LOS) and mortality. MAIN RESULTS: Full analysis was obtained for 80 patients. On hospital discharge, the mean increase in reticulocyte count was significantly higher in the O(2) group than the air group. Percent variation also increased: 184.9% ± 41.4% vs 104.7% ± 32.6%, respectively; P < 0.001. No difference in Hb or Hct levels was noted at discharge. Allogeneic red blood cell transfusion was 7.5% in the O(2) group versus 35% in the air group (P = 0.0052). Hospital length of stay (LOS) was significantly shorter in the O(2) group than the air group (7.2 ± 0.7 days vs 7.8 ± 1.6 days, respectively; P < 0.05). CONCLUSIONS: Transient O(2) administration increases reticulocyte count after traumatic hip surgery. Hospital LOS also was shorter in the O(2) group than the control group. Allogeneic red blood cell transfusion was reduced in the O(2) group but it was not due to the NOP mechanism.