Abstract: Abstract: The LUCC model is an important way to understand the process of land-use change, driving mechanisms, dynamic changes, ecological effects, and environmental impact assessment. Studies of land use models on land use prediction in mining cities are relatively rare. The CLUE-S model, as the representative of experience-based statistical models, has high simulation accuracy and application value in land use change prediction from spatial and temporal aspects. In this paper, the CLUE-S model was applied to land use change simulation in Wu'an, a typical mining city relying on GIS technology. First, selecting the correct driving factors is necessary to the accuracy of prediction map. 15 driving factors of land use were selected from 28 driving factors according to RDA(redundancy analysis) and factor analysis. Using 15 driving factors not only reduces the complexity of the problem, but also preserves simulation accuracy. In consideration of sustainable development, the free market scenario is more suitable compared with the alternatives. The expansion trend of the free-market mode is towards east and southeast, which conforms to the development planning of Wu'an city. Second, we set the corresponding land use quantity change in 2020 under three development modes by a Markov model and GM (1, 1) grey model, and then predicted the land use distribution map under a free market scenario, a cultivated land protection scenario, and an ecological protection scenario. The result showed that forest land and construction land increased under a free market scenario, reflecting the effects of economic development and environmental protection.. In the cultivated land protection scenario, cultivated land distribution restricted the free expansion of construction land. In the ecological security scenario, forestland grows obviously, and mining land reduces sharply influenced by environmental policy. Comprehensive consideration of sustainable development in Wu′an from the perspective of social, economic, ecological development and cultivated land protection, the free market scenario is relatively reasonable, and the other two scenarios can be a supplement for regional land optimal allocation. The results further verify that the CLUE-S model can simulate the future land use change of mining cities under different scenarios. Meanwhile, a new method to predict future land use under different scenario using the CLUE-S model can be applied in the implementation and management of land use planning, which can guide the land use change in the implementation process with reference to different simulation results by adjusting the land use objective and ultimately achieve the purpose of land use optimization.