Evapotranspiration is one of the most important parts of water cycle in the nature but, the measurement of actual evapotranspiration (via Lysimeter as an exact measurement instrument) is so difficult and not practical. Therefore, equations that can estimate the value of evapotranspiration only by using meteorological data are necessary. In this study, four different reference crop evapotranspiration of Penman-Montith FAO56 (PMF56), Penman-Montith FAO with Irmak Radiation (PMFI), Jensen-Hais1 (JH1) and Jensen-Hais2 (JH2) have been assessed in a semi-arid climate of Iran to estimate the ETo. The investigations showed that, when the data of radiation, temperature and wind changed by ±5% and ±10%, the radiation was more effective than other parameters. Therefore to calculate the short wave radiation, radiation models of Angstrom, Sabbagh, Gelever-Mc Culoch and Black have been used. Calibration of obtained results with data of three weighting microlysimeter during the growing season (May to October) duration 95 daysby using different statistical parameter such as coefficient of determination (R2), Root mean square error (RMSE), Mean bios error (MBE) and t-test criteria of 0.781, 0.625, 0.038 and 0.468 respectively, indicated that PMF56 model with radiation models of Angstrom is the best model for estimation of reference evapotranspiration and the models of JH1 and PMFI have low exactness for this climate. The finding of the present research reveal that in all investigated evapotranspiration models, the models that used Angstrom radiation have high accuracy and radiation models of Sabbagh, Gelever-Mc Culoch and Black are in next orders.