Plateau zokor (Eospalax baileyi) is a subterranean rodent endemic to the Tibetan Plateau, and it plays an important role in the alpine rangeland system because of the ecosystem function in terms of substance circulation and energy exchanges. However, at high population density, zokors are regarded as pests aggravating the degradation of rangelands. As zokors live underground, they dig tunnels in the soil for foraging or mating. In digging tunnels, the soil is pushed out on the ground as mounds, which reduce the grazing area and cover the forages. However, the mounds made by zokors create environmental heterogeneity, which increases the species diversity and nutrient recycling in soil. Therefore, the mounds have an important impact on the alpine rangeland ecosystem. Studying the spatial distribution of plateau zokor mounds and its correlation with environmental factors can reveal its role in habitat use and selection, which is the basis for controlling zokor damage and protecting biodiversity in the place. The variables of traditional statistics must be random variables, and the traditional statistics can not clarify the relationship between the variables and geographic location. As a result, the traditional statistics can not explain the reasons that influence the spatial distribution patterns in population studies. Currently, the geostatistical analysis methods are often used to study population spatial distribution patterns. In our study, we selected the small habitat of plateau zokor (140m×100m) in eastern Qilian Mountain because a small habitat can eliminate the heterogeneity of climate and geography. We used the semivariogram and ordinary kriging of geostatistical analysis methods to analyze the spatial heterogeneity of zokor mounds and the environmental factors that influence the distribution of mounds. The purpose of this study was to determine the relation between the distribution of zokor mounds and environmental factors, including soil bulk density, soil moisture, aboveground biomass, underground biomass, root's water-soluble sugar, crude protein, crude fat, richness of grass, and richness of forbs and sedges. The results of semivariogram and ordinary kriging indicated that the mounds of plateau zokor presented clumped distribution with medium variation level, and all environmental factors had spatial heterogeneity in the studying site. The results of a crossvariogram analysis showed that the distribution of zokor mounds had a positive spatial correlation with soil moisture and aboveground as well as underground biomass at the same studying scale (10.74-71.85m). The distribution of zokor mounds had a positive spatial correlation with forage nutrients, such as root's crude protein and water-soluble sugar at the studying scale (10.74-30.82m), as well as root's crude fat at the studying scale (10.74-44.01m). Besides, the distribution of zokor mounds had a positive spatial correlation with forbs at all studying scales. The mantel test indicated that the distribution of zokor mounds had a significant negative spatial correlation with soil bulk density and richness of sedge and a significant positive spatial correlation with richness of forbs and the contents of root's fat. In conclusion, the plateau zokors prefer to select the alpine meadows with softer soil, more forbs as well as plants with higher root's fat in this studying site.