Lake is one of the important surface water resources in arid regions. They not only play an important role in maintaining the regional ecosystem balance, but also as a sensitive indicator of climate and environmental changes. Mapping and monitoring the lake area changes are essential to formulate the regulation policy on local water resources utilization. According to the second National Lake Survey (2007-2011), there are 108 lakes (more than 1km²) in Xinjiang Uygur Autonomous Region. Compared with the first National Lake Survey (1960s-1980s), there are 62 Lakes (more than 1km²) disappeared in Xinjiang, which accounting for 87% of the total area of disappeared lakes in China. Disappeared lakes are the main sources of saline dust storm which become a serious threat to regional ecological security and human health. Spectral indices method was mostly conceived for water body identification on imagery from medium resolution, broad-band multispectral sensors like Thematic Mapper (TM), Enhanced Thematic Mapper Plus (ETM+), SPOT-5, etc. Thus, this study proposed a water body extraction method based on dynamic threshold of multiple remote sensing spectral indices. Based on the remote sensing cloud computing platform-Google Earth Engine (GEE), this study has analyzed the spatio-temporal characteristics of the lake numbers and areas in Xinjiang from 1986 to 2019 based on Landsat images. The main reasons for the lake changes were preliminarily discussed in aspects such as climate change and human activities. The results were as follows:1) The water body extraction method based on multiple spectral indices proposed in this study has a better identification result in different types of lakes in arid region. 2) The number of lakes in Xinjiang (more than 0.5km²) has been increasing overall since 1986. Especially in the past 10 years, the lake number has shown a sharp increase, mainly in the area around Tianshan Mountains. And in the northern part of the Qinghai-Tibet Plateau, the number of lakes (less than 5km²) has increased significantly. 3) During the study period, the area of main lakes (more than 10km²) showed a significant expansion trend, and lakes with significant expansion were mainly concentrated in the northern part of the Qinghai-Tibet Plateau. Ayakum Lake had the most significant expansion trend (+16.52km²/a., P<0.01) in the study area. However, lakes around Tianshan Mountains have a frequent inter-annual fluctuation in recent years. 4) Under the background of global warming, plateau lakes continued to expand steadily after 1997 due to the supply of abundant snow and ice melt water and minor human disturbances. On the contrary, the lakes around the Tianshan Mountain were simultaneously affected by accelerated retreat of glaciers and human activities. Therefore, the lakes have been fluctuating violently in recent years.