BACKGROUND: Perioperative anemia leads to increased morbidity and mortality and potentially inhibits rehabilitation after hip fracture surgery. As such, the optimum transfusion threshold after hip fracture surgery is unknown. PATIENTS AND METHODS: A total of 120 elderly, cognitively intact hip fracture patients admitted from their own home were randomly assigned to receive transfusion at a hemoglobin threshold of 10.0 g per dL (liberal) versus 8.0 g per dL (restrictive) in the entire perioperative period. Patients were treated according to a well-defined multimodal rehabilitation program. Primary outcome was postoperative functional mobility measured with the cumulated ambulation score (CAS). RESULTS: Patients in the liberal group received transfusions more frequently than those in the restrictive group (44 patients vs. 22 patients; p < 0.01) and received more transfusions during hospitalization (median, 2 units [interquartile range, 1-2] vs. 1 [1-2]; p < 0.0001). There were no significant differences in postoperative rehabilitation scores (CAS: median, 9 [9-15] vs. 9 [9-13.5]; p = 0.46) or in length of stay (median, 18 days vs. 16 days, respectively; p = 0.46). There were fewer patients in the liberal transfusion group with cardiovascular complications (2% vs. 10%; p = 0.05) and a lower mortality (0% vs. 8%; p = 0.02). CONCLUSION: Although a liberal transfusion trigger did not result in increased ambulation scores, restrictive transfusion thresholds should be treated with caution in elderly high-risk hip fracture patients, until their safety has been proved in larger randomized studies.