Purpose of the study After total hip replacement, the initial stability of the cementless femoral stem is a prerequisite for ensuring bone ingrowth and therefore long term fixation of the stem. For custom made implants, long term success of the replacement has been associated with reconstruction of the offset, antero/retro version of the neck orientation and its varus/valgus orientation angle. The goals of this study were to analyze the effects of the extra-medullary parameters on the stability of a noncemented stem after a total hip replacement, and to evaluate the change of stress transfer. Material and methods The geometry of a femur was reconstructed from CT-scanner data to obtain a three-dimensional model with distribution of bone density. The intra-medullary shape of the stem was based on the CT-scanner. Seven extra-medullary stem designs were compared: 1) Anatomical case based on the reconstruction of the femoral head position from the CT data; 2) Retroverted case of -15 degrees with respect to the anatomical reconstruction; 3) Anteverted case with an excessive anteversion angle of +15 degrees with respect to the anatomical case; 4) Medial case: shortened femoral neck length (-10 mm) inducing a medial shift of the femoral head offset; 5) Lateral case: elongated femoral neck length (+10 mm) inducing lateral shift of the femoral head offset 6) Varus case with CCD angle 127 degrees; 7) Valgus case with CCD angle 143 degrees. The plasma sprayed stem surface was modeled with a frictional contact between bone and implant (friction coefficient: 0.6). The loading condition corresponding to the single limb stance phase during the gait cycle was used for all cases. Applied loads included major muscular forces (gluteus maximus, gluteus medius, psoas). Results Micromotions (debonding and slipping) of the stems relative to the femur and interfacial stresses (pressure and friction) were different according to the extra-medullary parameters. However, the locations of peak stresses and micromotions were not modified. The highest micromotions and stresses corresponded to the lateral situation and to the anteverted case (micro-slipping and pressure were increased up to 35 p.100). High peak pressure was observed for all designs, ranging from anatomical case (34 MPa) to anteverted case (44 MPa). The peak stresses and micromotions were minimal for the anatomical case. The maximal micro-debonding was not significantly modified by the extra-medullary design of the femoral stem. Discussion The extra-medullary stem design has been shown to affect the primary stability of implant and the stress transfer after THR. Most interfacial regions present small micro-slipping which normally allows the occurrence of bone ingrowth. The anatomical design presents the lowest micromotions and the lowest interfacial stresses. The worst cases correspond to the anteverted and lateralized cases. Probably, the anteverted situation involves higher torsion torque, which in turn may induce high torsion shear micromotions and higher stress at the interface. Moreover, the lever arm of the weight bearing force on the femoral head is augmented for the augmented neck length situation. This increases the bending moment, and therefore may increase the stresses as well as the stem shear micromotions. In summary, the present results could be taken as biomechanical arguments for the requirement of anatomical reconstruction of not only the intra-medullary shape but also the extra-medullary parameters (reconstruction of the normal hip biomechanics).