Abstract: Surface evapotranspiration is an important part of terrestrial hydrological cycle, and the analysis of spatio-temporal variation characteristics of evapotranspiration is the basis of in-depth understanding of hydrological processes in arid areas. Due to the lack of long-term observations of actual regional evapotranspiration in the Yinchuan Plain, it is difficult to obtain the spatio-temporal variation characteristics of long-term series evapotranspiration. Based on the MOD16A3 surface evapotranspiration data and the measured data of meteorological stations in the study area, this paper analyzes the variation characteristics and influencing factors of surface evapotranspiration in the Yinchuan Plain from the perspective of time and space by using the methods of Theil Sen median trend analysis, MK mutation test and CA-Markov model, and predicts the development trend of surface evapotranspiration in the next five years. The results show that from 2004 to 2019, the interannual variation trend of evapotranspiration in the Yinchuan Plain is increasing, and the MK mutation test results show that the year of 2010 is the mutation point of evapotranspiration time series data. There are obvious differences in spatial distribution pattern and change trend between the actual evapotranspiration (ET) and potential evapotranspiration (PET) in the Yinchuan Plain. Actual evapotranspiration shows an increasing trend in recent 16 years, and potential evapotranspiration shows a decreasing trend, which conforms to the theory of complementary evapotranspiration in arid areas. The future development trend of surface evapotranspiration in the Yinchuan Plain in 2024 is predicted with the CA-Markov model. The simulation results show that the evapotranspiration will still increase in the next five years. The spatio-temporal variation of evapotranspiration is affected by climate and human activities. Evapotranspiration is positively correlated with temperature, precipitation and sunshine hours, and is negatively correlated with relative humidity. Land use structure affects annual evapotranspiration. The results also show the rule: Evapotranspiration of paddy field > dry land > woodland > grassland > desert.