Evaluating the efficiency of empirical methods for determining of reference evapotranspiration is essential in order to determine plants water requirement. Therefore, in this study, the meteorological data of 154 stations around the Iran country were used during the statistical period of 15 years (1999 to 2013). Potential evapotranspiration for all synoptic stations was computed using the computational methods based on air temperature, containing 17 methods. The results of different methods were compared to the result of the reference method (FAO-Penman-Monteith). The best and worst methods were selected based on different statistical indexes including regression coefficient, squared squares, mean error estimates and Jacobis criterion. The results showed that for dry climate, Hargreaves-M4, Hargreaves, Hargreaves-M2, Hargreaves-M1, Hammon 1961, Hamon 1963 and Hargreaves Samani were the most suitable methods, respectively.
For semi-arid climate, the Hargreaves-M4, Hargreaves, Hamon 1963 and Hargreaves Samani methods were the most appropriate methods. In the Mediterranean climate, the Hargreaves-M4 and Hargreaves methods were the most appropriate methods. In the semi-humid climate condition, the Hargreaves-M4 and Hargreaves-M2 were selected as the most appropriate methods. In wet climates, Hargreaves-M4, Hargreaves-M2 and Hargreaves-M1 methods were the most suitable methods and in the very humid climates, the Hargreaves-M4, Hargreaves, Hargreaves-M1and Blanicril-4 were the most appropriate methods respectively. Moreover, the Schendel method also was selected as the worst method in all the above mentioned climates. Meanwhile, in the semi-arid climate, the