Evaporation is an important component of hydrological cycle and key input in hydrological models. Quantification of spatial-temporal variation of potential evapotranspiration(PET)is helpful for better understanding climate change and its influence on hydrological cycle and water resources. In this study,the FAO Penman- Monteith was used to estimate the spatial-temporal variation of PET in the Loess Plateau with monthly observed data at 50 meteorological stations during 1959-2015 obtained from the National Meteorological Information Center of China Meteorological Administration. The Spline method was applied for interpolation to obtain the spatial distribution of PET. The sensitivity coefficients method was employed to explore the contribution of key meteorological variables to the variation of PET. The results show that mean annual PET of the Loess Plateau ranged from 780 mm to 1 470 mm and decreased from northwest to southeast. Seasonally,PET in summer,spring,autumn and winter decreased successively. During 1959-2015,PET showed an upward trend of 5.64 mm·(10 a)~(-1) in the Loess Plateau. PET showed the highest increasing trend in spring[7.37 mm·(10 a)~(-1)],followed by summer, autumn and winter. The west part,middle-north part and south-east part of the Loess Plateau showed nonsignificant increasing trends in PET. The increase of solar radiation is the dominant factor for the increase of PET,followed by actual water vapor pressure,wind speed and temperature. The innovation of this study is using the sensitivity coefficients method to analyze the effects of different meteorological variables on potential evapotranspiration. This study analyzed the spatiotemporal changes of regional PET and the influence of each meteorological variable. The results are favorable to the water resources planning and management,and provide scientific basis for the efficient agricultural water resources exploitation.