OBJECTIVE The determination of rotation of the femur during intramedullary nailing procedures can be difficult, particularly when the fracture pattern does not lend itself to interdigitation. We studied 22 patients who had isolated femur fractures treated by closed intramedullary nailing to determine the degree of malrotation introduced at the time of surgery. DESIGN AND METHODS Anteversion of the affected and normal femora was determined by a standard computed tomography (CT) torsion study. The range of motion of the hip was measured in the prone position. Foot progression angles (FPAs) were measured in 14 patients who were fully ambulatory for at least 6 months. MAIN RESULTS The average malrotation of the fractured femur was 16 degrees (4 to 61 degrees). The median malrotation was 14 degrees. The differences in CT-measured anteversion (delta A), FPA (delta FPA), internal rotation (delta IR), and external rotation (delta ER) between the affected and normal sides were determined. Linear regression was used to analyze delta A with delta FPA, delta IR, and delta ER. Changes in internal and external rotation as determined by physical exam had a stronger correlation with delta A than did delta FPA. This indicates that malrotation of the femur is accommodated for during gait. CONCLUSIONS Based on this data, we found that anteversion of the normal femur can be determined in the operating room using the image intensifier and can be duplicated on the fractured side using the described technique in cases where comminution prevents fragmentational alignment. This method has been used for 12 patients in a prospective trial, and malrotation has been kept to under 10 degrees in all cases.