In the recent decades, increased climate change and human activities have been affecting the spatial and temporal distribution of water resources and water balance components in the inland region of Northwest China. This study used a remote sensing cloud computing platform, Google Earth Engine (GEE), to explore the spatial and temporal dynamics of water balance components in this area at both basin and sub-basin scales with both interannual and intraannual analyses over the last two decades. Multi-source remote sensing datasets were employed to fulfill this purpose. The study tried to reveal the spatial distribution characteristics and temporal change of key water balance components, including terrestrial water storage, precipitation, soil moisture, evapotranspiration, and surface water. The results show that: (1) At the whole basin scale, due to the increasing trend of air temperature, glaciers and snow in high altitude areas were melting with an increasing speed, which provides more water supply to the low altitude basins. Therefore, the terrestrial water storage exhibit a pattern of “decreasing at high altitude and increasing at low altitude”. Meanwhile, surface water area, soil moisture and evapotranspiration also showed a certain increasing trend accordingly. (2) At temporal scale, except for the desert areas, different water balance components exhibited different seasonal variations. The value of each component generally reaches the maximum in the summer and the autumn, and becomes the lowest in the spring and the winter. (3) At sub-basin scale, due to the different combinations of water sources, the variation patterns between different water balance components differ from sub-basin to sub-basin. In the ice and snow water dominated sub-basins, each component showed a similar variation pattern as the terrestrial water storage, while in the precipitation dominated sub-basins, the variation of each component is more closely related to precipitation fluctuation. In conclusion, different water balance components exhibit different characteristics at different spatial and temporal scales in the inland area. This study provides valuable reference for further understanding the co-evolution pattern of these components.