Abstract: In the Manas River Basin the climate is dry and evaporation is strong, and accurate estimation of evapotranspiration is of important guiding significance for groundwater resources evaluation and ecological environmental protection. In the past, the research on evapotranspiration is of a low spatial resolution and the results cannot satisfy the detailed changes in evapotranspiration caused by the evolution of landscape patterns in various hydrogeological zones. In response to the past deficiencies, this study uses the Landsat series of images to estimate the evapotranspiration in the piedmont plain of the Manas Basin in the past 30 years, and further explores the temporal and spatial distribution characteristics and influencing factors of evapotranspiration in different hydrogeological zones. The results show that the spatial distribution of evapotranspiration is obviously banded according to the hydrogeological division. The total evapotranspiration of each hydrogeological zone is represented by the Gobi belt < desert belt < oasis belt. The total evapotranspiration of the whole area shows an increasing trend on the time scale and the increasing trend gradually slows down. The total amount of evapotranspiration in the oasis belt increases, and the total evapotranspiration in the gobi belt and desert belt decreases first and then increases. The change trend of evapotranspiration in each district is controlled by the variation in evapotranspiration of the major land types in each district. According to the analysis of influencing factors, it can be found that ET increases with the rising temperature, and the daily average evapotranspiration of various places is consistent with the average temperature change trend of the whole region. NDVI and ET show a good positive correlation in the gobi and oasis zones. Groundwater table burial depth is negatively correlated with ET in the oasis zone. When the groundwater table burial depth is greater than 5.5 m, ET tends to stabilize. The analytical ideas can provide reference for hydrogeologists in the data-poor areas.