Abstract: Reference crop evapotranspiration (ET₀) can fully represent the evapotranspiration capacity in areas, such as efficient water-saving irrigation in agriculture. Most studies have typically conducted sensitivity analyses of changing ET₀ in the local meteorological factors, without considering the remotely correlated effects of large-scale climate variability on ET₀. Large-scale climate variability generated by the ocean is a significant driver of global and regional climate change, indicating the non-negligible impact on the evolution pattern of ET₀. However, it is still lacking in the influence of large-scale climate variability on ET₀, leading to the inaccurate attribution of ET₀ changes. Therefore, this study aims to explore the relationships between the large-scale climate variability of El Niño Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), Pacific Decadal Oscillation (PDO), and Atlantic Multidecadal Oscillation (AMO) with the ET₀ variability characteristics in Xinjiang region. The daily meteorological data and climate variability indices were collected from 84 meteorological stations using multiple linear regression and Cramer's mutation test. The results indicate that the ET₀ showed a decreasing trend from 1960 to 2020, with an average decreasing rate of 0.75 mm/a; the year 1998 was the mutation of the ET₀, and the ET₀ showed an obvious decreasing trend from 1960 to 1997, with an average decreasing rate of 2.50 mm/a, and then turned to an outstandingly increasing trend from 1998 to 2020, with an average increasing rate of 3.18 mm/a. The multivariate linear correlation results show that the ET₀ was negatively correlated with the RH with a regression coefficients of −0.46, and positively correlated with the U, T, and R_n with regression coefficients of 0.95, 0.36, and 0.20, where the changes of ET₀ were mainly dominated by the changes of U. On the seasonal scale, the interannual variation trend of ET₀ in the four seasons was more consistent with the annual cumulative ET₀ variation trend, which decreased at the rate of 0.01, 0.54, 0.23, and 0.01 mm/a in spring, summer, autumn, and winter, respectively, whereas, the trend change in the spring, summer, and autumn seasons occurred in the mid-1990s, respectively. PDO was the main variable affecting the trend shift of ET₀, where the regression coefficients between the two was -0.34. There was also a significant negative correlation of U with PDO and AMO, with regression coefficients of −0.31 and −0.34, respectively. The PDO was shifted from a positive to a negative phase in 1998, resulting in the wind speed shifting from a downward to an upward trend, which in turn led to the trend shift of ET₀ in the Xinjiang region in 1998. The shift of AMO from a negative to a positive phase strengthened the trend change of wind speed, while the effects of ENSO and IOD on wind speed were relatively weak. The negative phase of PDO was achieved in the spatial regression of sea surface temperature (SST) with ET₀, which further verified the influence of PDO on the trend shift of ET₀ in the Xinjiang region of western China.