Grassland degradation is one of the important environment and resource issues in North China. Different degrees of grassland degradation in Hulunber grassland are mainly caused by global climate change, overexploitation and other reasons, which threats to the sustainable development of grassland animal husbandry and improvement of living standards of farmers. The relationship between spatial patterns and ecological processes at different scales is one of the hottest research topics, the key point of which is to select an appropriate observation scale. Disturbance as part of the process, is the main sources of spatial heterogeneity, which can significantly change the ecological pattern and processes at different scales, causing apparently shift of the structure and function of ecological systems. Small-scale spatial structure of plant species was important in plant ecology as it affected the dynamics of plant populations, communities and ecosystem processes. Hulunber steppe with a plateau landscape is a component of the Mongolian Plateau in central Asia and located in the eastern part of Inner Mongolian, which with the unique ecosystem characteristics, relatively well-preserved natural environment, abundant landscape types and biodiversity, relatively advanced and intensive production mode, is the optimal object for researching the natural ecological patterns and process in steppe ecosystems, and its responses to human-caused disturbances. Leymus chinensis, the zonal vegetation in Hulunber grassland, has a great significance for understanding ecological processes and sustainable utilization of grassland resources with studies on population patterns of Leymus chinensis. We recorded fine-scale (0-100cm) spatial patterns of this species as presence/absence maps from 2m×2m grids (2cm×2cm quadrats) in two plots under three kinds of disturbances: cutting, fencing and grazing. Ripley′s K function and Monte Carlo stochastic simulation method were used to quantitatively analyze Lemus chinensis population patterns. Based on life history characteristics, strategies of ecological adaptation of Leymus chinensis population and interspecific interaction in the community under different disturbances, the mechanism generating and maintaining these patterns was discussed. The results showed that: First, there were significant effects on spatial patterns of Lemus chinensis populations under the disturbances of cutting, fencing and grazing in the Restoration Succession Sample of Leymus chinensis Meadow Steppe(abbr. LM) and Long-term Observation Sample of Stipa baicalensis Meadow Steppe(abbr. SM). The density of Leymus chinensis population in LM was significantly greater than SM, but its L(h) value in LM was lower than SM, which indicated the aggregation level of SM was greater than LM. Second, there is a difference at virous scales. On the scales less than certain critical points with the population usually characterized by non-random distribution, is being regular in LM and clumped in SM respectively; In contrast, if the scales enlarge beyond these critical points, the plants will disperse randomly but without disturbances of cutting which is clumped. It is mainly related to Leymus chinensis′s reproductive responses community′s position, and competition ability for resources under different disturbances. Third, the nature of spatial structure seemingly depends on the scale of observation, but the smaller-scale data are likely to provide a more powerful biologically explanation for local spatial structure in this community.