Lake water resources are vital in Xinjiang, northwestern China, a mountainous region that depends highly on melting runoff, making it susceptible to climate change. The interannual dynamics of lakes in Xinjiang and the driving mechanism of climate factors are however poorly understood. Here, we used the newly developed monthly water map datasets, the climate dataset from the ERA5 reanalysis, hydrological datasets from Gravity Recovery and Climate Experiment (GRACE) and Global Land Data Assimilation System (GLDAS), the Pearson correlation analysis, and a geographically and temporally weighted regression to characterize the spatial-temporal dynamics of the lakes in Xinjiang from 2000 to 2019 and further to explore their response to climate factors. We found that the number of lakes and the area that they cover in Xinjiang have increased significantly over the past two decades, with average increase rates of 5 lakes/year and 63.26 km²/year, respectively. The most significant expansion of lakes occurred in the alpine areas (3000–5000 m). Precipitation can directly contribute to lake expansion for almost all lakes. Influences of recent warming were more complicated, with the promotion effects occurred in higher bands. We found lakes in Tianshan Mountains, fed by the most rapidly retreating glaciers, were relatively stable or even contracted. Remarkable lake expansion was found in the continuous permafrost zone of Kunlun Mountains, where the climate warming resulted in an increase in meltwater from ground ice, which has contributed to increasing of ground water storage and accelerated lake expansion. The overall expansion of lakes in Xinjiang is alarming because an initial short-term increase may be followed by a long-term shrinkage due to the increasing consumption and retreat of ice materials. This research provides data to support water resources management in Xinjiang and is also relevant for climate change in other regions that receive water from surrounding mountains.