Evapotranspiration (ET) is a critical component of the land-atmosphere water cycle in the natural environment and a significant element in surface energy balance processes. This study presents an enhanced version of the surface energy balance system (SEBS) model, termed the modified SEBS (mSEBS) model, to estimate the daily actual land surface ET in China, with a spatial resolution of 500 m, from 2011 to 2022. The improvements were implemented for two key aspects: strengthening the input data and refining the model structure. Model validation via data from 32 flux tower sites revealed a slight overestimation of ET, with a mean bias error (MBE) of 0.07 mm/day. Nevertheless, the overall model performance was robust, as demonstrated by a high correlation coefficient (R) of 0.80 and a relatively small root mean square error (RMSE) of 0.93 mm/day. Further validation against water balance-derived ET (WB_ET) data from China's nine major river basins also yielded impressive results, with a high correlation (R = 0.90) and minimal annual bias (103.70 mm). Site-scale validation across various underlying surfaces, landforms, and climatic regions indicated that the model performed best for grasslands, plains, and semiarid regions, respectively. Spatiotemporal analysis revealed a significant decreasing trend in ET (-1.76 mm/year, p < 0.05) only in the continental river basin (CORB) over the 12-year study period, whereas a nonsignificant decreasing trend (-0.72 mm/year, p > 0.05) was observed at the national scale. This study aims to provide a scientific basis for water resource management and allocation while contributing to research on climate change, sustainable development, and food security.