Different hypotheses (geographic, ecological, evolutionary or a combination of them) have been suggested to account for the spatial variation in species richness. However, the relative importance of environment and human impacts in explaining these patterns, either globally or at the biogeographic region level, remains largely unexplored. Here we jointly evaluate how current environmental conditions and human impacts shape global and regional gradients of species richness in terrestrial mammals. We processed IUCN global distributional data for 3939 mammal species and a set of seven environmental and two human impact variables at a spatial resolution of 96.5x 96.5 km. We used simple, multiple and partial regression techniques to evaluate environmental and human effects on species richness. Actual evapotranspiration is the main driver of mammal species richness globally. Together with our results at the biogeographic realm level, this lends strong support for the Hawkins et al. (2003) conjecture (i.e. global diversity gradients are best explained by the interaction of water and energy, with a latitudinal shift in the relative importance of ambient energy vs. water as we move from the poles to the equator). While human effects on species richness are not easily detected at a global scale due to the large proportion of shared variance with the environment, these effects significantly emerge at the regional level. In the Nearctic, Palearctic and Oriental regions, the independent contribution of human impacts is almost as important as current environmental conditions in explaining richness patterns. The intersection of human impacts with climate drives the geographic variation in mammal species richness in the Palearctic, Nearctic and Oriental regions. Using a human accessibility variable we show, for the first time, that the zones most accessible to humans are often those where we find lower mammal species richness.