Land surface evapotranspiration (ET) and its partitioning between evaporation (E) and transpiration (T) is a significant component of water and energy cycles at all scales, from field and watershed to regional and global, and is essential to many applications in climate, weather, hydrology, and ecology. The land surface ET and its components E and T can be produced conveniently at a range of spatial and temporal scales by combining the advanced remotely sensed data and its land surface products such as land surface temperature, leaf area index, and landcover, among others. This work aims to evaluate and summarize available remotely sensed models currently used to determine ET and components E and T. The remotely sensed-based model of land surface E and T has undergone several stages of development, including series and parallel energy balance models, spatial variability model, remote sensing and meteorological combination model, and data assimilation technology divided based on diverse model mechanisms. However, these models provide wide ranges of E and T, whose uncertainty may be limited by the unreasonable component temperatures partitioned from land surface temperature, parameterization of the stress factors of T from vegetation and E from soil surface, and uncertainty of the reproduced meteorological data as model input data. Future studies should improve model performance under heterogeneous surface and upscale the point or patch ground measurements of E and T to satellite pixel scale to validate remotely sensed model simulations.