The relationship of vegetation with climate change has been identified in recent studies at regional and global scales. These efforts have provided the basis for vegetation reconstruction and eco-environment restoration through research on vegetation dynamics in specific regions and on its relationship with climate change to find the key factor affecting vegetation changes. In this study,the simulations of climate (mainly temperature and precipitation distribution and changes over Xinjiang) in the Xinjiang region were carried out by RegCM3 nested into CRU data. The RegCM3 model has the ability to simulate the basic character of the spatial distribution of precipitation and temperature over Xinjiang. At the same time,the model simulation can provide more reliable information of local precipitation and air temperature for regions with few observation stations. The impact of the change in rangeland vegetation type on climate in Tianshan Mountains was analyzed by the Regional Climate Model. Using this model, we analysed the impact of vegetation change on temperature and precipitation through three sensitivity experiments. The increase in grassland area may have little effect on temperature and precipitation in the Tianshan Mountains, but vegetation type that changed from grassland to cropland resulted in an increase of temperature and precipitation in the surface layer. Half of the grass area cultivated could bring growth from May to July in the Tianshan Mountains. This study showed that latent heat release was the main factor causing the ground temperature difference. The latent heat, by influencing the water vapor contents, affected the thermo-convective local precipitation.