The Lake of Valencia basin, Venezuela, covers an approximate area of 3150 km2 and within it, there is only 1 rainfall station that has good quality in the historical record of meteorological data. Because of the lack of accurate meteorological information, monitoring of evapotranspiration in the Lake of Valencia basin is currently deficient. Due to the importance of the estimation of the evapotranspiration for the management of water resources, have been developed numerous direct and indirect methods based on meteorological data measured in situ. Although they have demonstrated the feasibility of their use, the estimation of the evapotranspiration using these methods presents great limitations in wide areas by cause of the heterogeneity of the surface and the dynamics of the heat transfer processes. Remote sensing techniques using multispectral images are currently a promising and low-cost alternative for estimating evapotranspiration at a regional scale with acceptable degrees of accuracy. In the study case, an indirect method was developed for the estimation of the evapotranspiration, based on the multispectral images of the Landsat 8 OLI sensor. The spatial resolution of the Landsat 8 products represents a technological advantage for the study of the regional variation of the evapotranspiration in the heterogeneous surface and vegetation conditions of the Lake of Valencia basin. The obtained models estimated 0.99, 1.01 and 1.05 of the average annual evapotranspiration calculated with the Hargreaves method for the years 2014, 2015 and 2016 respectively with a maximum percentage error of 5.19%. Among the main advantages of the developed models are the simplicity of use and spatial precision of 30 m, analogous to the spatial resolution of the OLI 8 sensor. © 2021, Univelt Inc. All rights reserved.