Abstract: Abstract: Land Use/Land Cover (LULC) has been one of the most important indicators to classify the human activities and natural elements on the landscape within a specific time. The complex types of various source materials can also involve representing the natural attributes and characteristics of the land surface. The spatiotemporal variation can be used to characterize the history of human social and economic development. There are also some influences on the global biochemical cycle, land-atmosphere hydrothermal cycle, and climate change. The Xilingol Steppe is located in the middle east of the Inner Mongolia Plateau, an important part of temperate grassland in Eurasia. However, the ecological environment of grassland is very fragile in recent years, due mainly to climate change, as well as unstable agricultural and animal husbandry production. Therefore, it is a high demand to scientifically reveal the temporal and spatial evolution characteristics and driving mechanism of LULC, in order to promote ecological civilization and high-quality development in the grassland. Taking the Xilingol Steppe as the study area, this study aims to identify the spatiotemporal characteristics of LULC in the recent two decades using the land use change index, frequency statistic, Sankey diagram visualization model, and grid cell. A GeoDetector was also employed to analyze the impact mechanism of natural geographical conditions and socio-economic development. Firstly, five periods of GLC_FCS30 data were selected to calculate the areas of each category in each period, in order to reveal the basic components of LULC in the study area. The frequency statistic was also conducted to determine the spatial distribution. Secondly, the area change and annual change rate of LULC in each period were calculated to examine the changing pattern of each category. The Sankey diagram was then constructed to describe the conversion between different LULC types. Thirdly, the grid cell of rate was obtained for the variation in the LULC in each period for the spatial patterns of LULC change. Finally, the digital elevation model, meteorological and socio-economic data from 2000 to 2020 were combined as the driving factors to clarify the mechanism of LULC change using the GeoDetector. The results showed that: 1) The LULC types were mainly grassland and sandy land (accounting for nearly 98% in total) from 2000 to 2020. The regional LULC change occurred in the sand/grass ecotone in the middle-west areas, due to the natural environmental conditions. 2) The area of sandy land increased significantly from 2000 to 2005. Specifically, the area of forest land, grassland, wetland, and water area decreased during this time. Nevertheless, there was a decreasing trend in the area of the sandy land year by year after 2005. Furthermore, the rate of conversion was still close to 5% in the recent ten years, indicating the generally good grassland ecological environment. But, it was still a high level of ecological vulnerability so far. 3) The GeoDetector results showed that the rapid growth of the social economy was the main driving factor of LULC change, followed by the climate factors, while the topographic condition was the relatively small influence factor. In general, the ecological environment was greatly improved over the last 15 years, although the relevant ecological restoration still needs to be further implemented. The findings can provide a scientific basis for ecological protection and sustainable development in the Xilingol Steppe. The approaches here are also applicable to the LULC change research in other similar regions.