This work suggests a densitometric protocol for femoral studies in arthrosic patients candidate to total hip arthroplasty. The protocol was applied retrospectively to 60 cases to investigate its possible future use as a tool for the preoperative planning of prosthetic stem fixation techniques-i.e., cemented vs. uncemented. The well-known standard procedures for femoral densitometric studies by dual-energy X-ray absorptiometry (DXA), are usually applied to the femoral neck, Ward's triangle and the greater trochanter. Unfortunately, these analyses lose their significance if applied to arthrosic hips, and arthrosis is the main condition requiring total hip arthroplasty. Our protocol for femoral studies in arthrosic patients includes the preoperative analysis of 7 regions of interest in the proximal shaft, similar to Gruen's regions. In vivo precision tests show variation coefficients ranging 0.9-3.7%. The protocol was applied to a series of 60 arthrosic patients who were submitted to total hip arthroplasty by the same surgical staff. Twenty of them were implanted a cemented stem and the other 40 an uncemented ("press-fit") stem. Densitometric results were communicated only after surgery, to avoid any bias on the normal decision-making process usually used to choose the fixation technique for the prosthetic stem. The two groups of cemented vs. uncemented stems had different distribution of densitometric values. Therefore, the information obtained with femoral densitometric studies preoperatively can play a major role in the choice between cemented and uncemented stems. Moreover, if femoral densitometric studies are carried out preoperatively, their results can be closely related to periprosthetic tissue remodeling and implant follow-up.