Preclinical wear evaluation is extremely important in hip replacements, wear being one of the main causes of failure. Experimental tests are attractive but highly cost demanding; thus predictive models have been proposed in the literature, mainly based on finite element simulations. In such simulations, the effect of friction is usually disregarded, as it is considered not to affect the contact pressure distribution. However, a frictional contact could also result in a shift of the location of the nominal contact area, which can thus modify the wear maps. The aim of this study is to investigate this effect in wear prediction for metal-on-metal implants. Wear assessment was based on a purpose-developed mathematical model, extension of a previous one proposed by the same authors for metal-on-plastic implants. The innovative aspect of the present study consists in the implementation of a modified location of the nominal contact point due to friction, which takes advantage of the analytical formulation of the wear model. Simulations were carried out aimed at comparing total and resurfacing hip replacements under several gait conditions. The results highlighted that the adoption of a frictional contact yields lower linear wear rates and wider worn areas, while for the adopted friction coefficient (), the total wear volume remains almost unchanged. The comparison between total and resurfacing replacements showed higher scaled wear volumes (wear volume divided by wear factor) for the latter, in agreement with the literature. The effect of the boundary conditions (in vivo versus in vitro) was also investigated remarking their influence on implant wear and the need to apply more physiological-like conditions in hip simulators. In conclusion although friction is usually neglected in numerical wear predictions, as it does not affect markedly the contact pressure distribution, its effect in the location of the theoretical contact point was observed to influence wear maps. This achievement could be useful for increasing the correlation between numerical and experimental simulations, usually based on the total wear volume. In order to improve the model reliability, future studies will be devoted to implement the geometry update by combining the present model to finite element analyses. On the other hand, further experimental investigations are required to get out from the wide dispersion of wear factors reported in the literature.