Abstract: Although several empirical or semi-empirical equations have been proposed to estimate the reference evapotranspiration (ET0), most of them were developed under specific conditions and, generally, cannot be applied under situations different from those under which they were originally developed. The applications of these equations in China should be studied. ET0 was calculated using ten different equations and data were collected from the Loess Plateau. The Penman-Monteith equation was used as a standard to be compared with the other equations for estimating ET0. The results of linear regression, using root mean squared error and mean bias error approaches show that there are good correlations among Penman type equations and that the best relationship was obtained between the Penman-Monteith and Kimberly Penman-72. The differences between those equations are larger in summer than that in spring and autumn. The Priestley-Taylor equation gave the best ET0 estimate among the simple equations, while the FAO-Rad, FAO Blaney-Criddle, Hargreaves and Makkink methods either overestimated or underestimated ET0 when compared with Penman-Monteith. Differences among those methods were also due to dependence on location. Therefore, the empirical parameters of those equations should be changed before being used in given local stations.