At present the remote sensing technology has become one of the most essential tools in the landscape ecology research. This paper is based on RS images from Landsat TM (1990), ETM (1999) and Aster (2004) imagery, with the support of RS and GIS technology and utilization of landscape ecology principle. Selected suitable image processing system, classification approaches and techniques and made a quantitative analysis of the study area's grassland landscape pattern change. RS image processing software (PCI) and landscape pattern analyzing software (Fragstats) is used in this article. The result shows that: in 1990, 1999 and 2004 the land use landscape pattern in study area experienced the considerable change. As far as the dynamic change of grassland landscape in the study area is concerned, natural factor takes certain driving action but human activities are the main driving force.