“Alpine-oasis-desert” is a unique natural landscape in arid areas, its internally distinct ecosystems are prone to different fluctuations under global change. As an important indicator to evaluate the quality of the ecological environment, vegetation net primary productivity ( NPP ) is of great significance to the overall understanding of regional changes. The generation of remote sensing images allows large-scale and long-term regional NPP estimation. The maximum light energy utilization efficiency under different plant covers classified by land-use type data also improves the accuracy of NPP estimation. Therefore, this paper selects the Yarkand River Basin with a typical alpine-oasis-desert ecosystem as the study area, using remote sensing image data and meteorological data for many years, selecting the CASA model based on light utilization rate simulate and analyze the NPP status of each ecological area. The following conclusions were drawn: (1) The annual average value of NPP in the Yarkand River Basin showed a fluctuating upward trend after 2 000, and about 85.9 % of the regions showed an upward trend. In the water area and the residential location of the oasis area, the NPP decreased. (2) The variation of NPP in the basin strongly correlations with precipitation, and its spatial distribution characteristics have an opposite correlation with NPP and temperature. (3) NPP in the Yarkand River Basin showed the highest in the oasis, followed by the desert-oasis transition zone, and the lowest in the alpine and desert regions. The fluctuation of NPP in the regions with relatively more fragile ecosystems (desert and alpine) was more substantial than that in the oasis and desert-oasis transition zone. The research results will provide theoretical support for restoring regional ecological environment protection, the response to climate change, the coordinated development of human beings and nature, and the promotion of multi-ethnic common prosperity.