Environmental change will affect the sustainability of human-environmental systems with natural and anthropogenic driving forces, especially in the ecologically vulnerable dryland, where the loss of biological and economic productivity hinders the prospects of poverty reduction. To adapt to a complicated changing environment, human society has to respond with various adaptive management and human behavior adjustments. The environmental state variance is greatly dependent on the system vulnerability, which was defined as the balance between exposure to physical threats to human well-being and the capacity of people and communities to cope with those threats. Vulnerability research can help to identify coping strategies for dryland regions. The present study provides a vulnerability assessment for the semiarid grasslands of the Xilingol League, Northern China. The grassland cover change of Xilingol is computed using the AVHRR NDVI data for 2000, by comparing it to the NDVI background from 1981 to 2006. The driving forces of grassland cover change are investigated by correlation analysis between grassland cover change and its potential climatic, topographic and anthropogenic drivers. An combined exposure-sensitivity index was calculated across the region using Spatial Principal Component Analysis (SPCA) for climate and anthropogenic indicators that had a strong correlation with the observed grassland cover change. This indicator was compared with an adaptive capacity index, constructed using principal component analysis (PCA) for relevant variables from aspects of location, economic development level, natural resource availability and administrative efficiency. As the relationships between exposure, sensitivity and adaptive capacity are not clearly understood, A vulnerability map was constructed by classifying and overlaying the spatial distributions of exposure-sensitivity and adaptive capacity. The results show that in Xilingol, the exposure-sensitivity index increases in a radial pattern with the valley in the northwest and the adaptive capacity index decreases from northeast to the southwest; the two indices are negatively correlated with each other, showing that harsher environmental conditions leading to higher exposure-sensitivity which failed in supporting the socio-economic and ecological infrastructure require a greater adaptive capacity. Combining exposure-sensitivity and adaptive capacity, the northeast part of Xilingol is identified to be least vulnerable due to a more favorable resource status and greater economic development. By contrast, the counties in the southwest, with harsh environmental conditions and a poor socio-economic infrastructure, have the greatest vulnerability and are in dire need of targeted adaptation measures to avoid a further decline in human well-being. The vulnerability of Xilingol decreases from southwest to northeast, which is opposite to the trend of annual mean precipitation, which indicates that the drier areas are more vulnerable than others. Based on the combination of exposure-sensitivity and adaptive capacity, Xilingol can be divided into eight sub-regions and the results of SPCA and PCA help in pinpointing the source of exposure-sensitivity and the advantages of adaptive capacity, which gives clues to find out a specific way for each county to reduce vulnerability by alleviating exposure-sensitivity and improving adaptive capacity. This paper demonstrated a straightforward approach to regional vulnerability assessment that can be readily applied in other dryland regions.