Femoral neck fractures are an important clinical, social and economic problem. Even if many different attempts have been carried out to improve the accuracy predicting the fracture risk, it was demonstrated in retrospective studies that the standard clinical protocol achieves an accuracy of about 65%. A new procedure was developed including for the prediction not only bone mineral density but also geometric and femoral strength information and achieving an accuracy of about 80% in a previous retrospective study. Aim of the present work was to re-engineer research-based procedures and develop a real-time software for the prediction of the risk for femoral fracture. The result was efficient, repeatable and easy to use software for the evaluation of the femoral neck fracture risk to be inserted in the daily clinical practice providing a useful tool for the improvement of fracture prediction.