Estimating reference crop evapotranspiration (ET0) is crucial for improving water-use efficiency, and its accuracy is influenced by climatic variability. This study evaluated 26 empirical ET0 models against the FAO-PM across four areas of China: Hangzhou (HZ) in the Eastern, Harbin (HB) in the Northeast, Urumqi (UQ) in the Northwestern and Lanzhou (LZ) in the North-central. Long-term (1990-2023) climate trends were first examined to identify regional influences on ET0. Based on the K & ouml;ppen-Geiger classification, the HZ was humid, the HB was semihumid, and the UQ and LZ were arid to semi-arid. The Mann-Kendall analysis indicated increasing precipitation in the HZ, HB and LZ, and significant warming in the UQ and HZ regions. Temperature and solar radiation emerged as the dominant climatic drivers of ET0, particularly in arid regions, whereas relative humidity and precipitation exerted a suppressive effect in humid areas. Model evaluation using NRMSE, %BIAS, MD, KGE and the compromise programming index showed that radiation-based models consistently achieved the best performance. Temperature-based models performed well in arid regions, whereas mass-transfer-based models tended to underestimate ET0. The findings underscore the need to match ET0 estimation methods with regional climatic conditions to support efficient irrigation management.