Abstract：The paper studied the influence factors of habitat selection by the plateau zokor (Myospalax baileyi) in alpine meadow in eastern Qilian Mountain under five grazing intensity. Providing a scientific basis for control grassland pest reasonably and biodiversity protection. In the five different grazing intensity areas, including heavy grazing (Ⅰ), secondary heavy grazing (Ⅱ), moderate grazing (Ⅲ), secondary light grazing (Ⅳ), and light grazing (Ⅴ). And monitoring relative plateau zokor population density, vegetation and soil change for three consecutive years, with the increased number of new mounds per unit area for three years to represent the plateau zokor population density (Table 1).The investigation method of grassland vegatation depended on the book of grass work technical guide, vegetation coverage, height, frequency, aboveground biomass and plant roots’ biomass were measured. the relatively important value of plant species in the community based on the plants’ relative coverage, relative frequency and relative height, then according to the important value to calculate the species’ uniformity (Je), abundance (He′) and diversity index (λ). And got the soil compaction by the soil compactness meter (SC-900, USA Spectrum). The index of relative zokor population density, vegetation (coverage, height, frequency, biomass) and soil (compaction, bulk density, moisture) were measured. Using SPSS (Statistical Product and Service Solutions) software to analyze the correlation between the relative population density of zokor and the other indexs, and also analyzed the difference of all indexs in five test plots in addition to the relative population density of zokor. The results showed that the relative population density of the zokor was lowest under the moderate grazing in five grazing intensity, which means the rangeland was not the zokor favorite habitat. However, the relative population density of the zokor in light grazing and minor grazing rangeland were higher than that in heavy grazing and extreme grazing rangeland. The rangeland under light grazing fitted the zokor’ inhabitation (Fig. 1). There were the significant positive correlation between the relative population density of zokor and the root biomass (R = 0.982, P < 0.01, Fig. 2b), vegetation abundance (R = 0.921, P < 0.05, Fig. 2d), and vegetation coverage (R = 0.909, P < 0.05, Fig. 2f). Aboveground biomass (R = 0.841, Fig. 2a), evenness (R = 0.873, Fig. 2c) and diversity (R = 0.921, Fig. 2e) were positively correlated but no significant (P > 0.05). There were the significant negative correlation (P < 0.05) between the relative population density of zokor and the soil compaction (R = -0.921, Fig. 3a,), soil bulk density (R = -0.883, Fig. 3b). And the significant positive correlation (R = 0.879, P < 0.05, Fig. 3b) with the soil moisture. The date suggest that the soil compaction was maybe the key factor that influence the habitat selection of the plateau zokor and the food was a secondary factor in the grassland grazing systems.