Estimation of ET0 in high-elevation (above sea level, ASL) areas of Tibet presents unique challenges: scarcity of monitoring stations, short-time coverage of meteorological data, low-oxygen and low-pressure environment, and strong solar radiation. In this study, altitude factors and modified temperature constants are utilized to improve the Hargreaves (HS) model based on 30-year daily meteorological data from 20 typical sites in Tibet. The improved model, Hargreaves-Elevation (HS-E) improved model, could provide better results at different time scales. Negative ET0 values were unavoidable in the HS model when applied to high-elevation areas. However, the HS-E model solved this problem and improved the accuracy of estimated ET0. In particular, the HS-E model performs better when the time scale becomes larger. Therefore, the HS-E model is highly recommended to estimate ET0 in high-elevation areas where the meteorological data is scarce, for example, in Tibet above 2,000 m.