Geographical isolation is one of major driving factors in shaping species distribution patterns. This study used the geographical distribution data of large and medium-sized terrestrial mammals in China to detect bioregions based on the Infomap Bioregions Method. The boundary of bioregion may represent the isolation effect of large and medium-sized terrestrial mammal groups. We applied a spatial autoregressive model to explore the main driving factors related to biogeographic boundaries starting from the gradient change of temperature, precipitation, and terrain. We constructed a Geographically Weighted Logistic Regression (GWLR) model to analyze the non-stationary spatial characteristics of the importance of each driving factor. Three main bioregions were identified. In addition to the coefficient of variation (CV) of annual precipitation was not obvious. The significance levels of the other variables at or below 10% were positively correlated with the biogeographic boundaries, which were related to the synergistic effects of climate mutation, topographic obstacles, and human activities. The dominant influencing factor of the boundary between bioregions 1 and 3 was the human footprint index, and that between bioregions 1 and 2 was the CV of temperature seasonality standard deviation. The main influencing factor of the boundary between bioregions 2 and 3 in the Qilian Mountains was the CV of temperature seasonality standard deviation, and that of the other areas was topographic obstacles. The response of multi-taxon species to isolation barriers was difficult to unify. Starting from large and medium-sized terrestrial mammals, this study identified biogeographic regions characterized by different species assemblages, revealed the spatial grouping rules of species, and provided a spatial model and framework to discuss the isolation mechanisms of various animal taxa.