Evapotranspiration (ET) is an important component of the hydrologic cycle, especially for irrigated agriculture. Direct methods of estimating reference ET are difficult or require many weather variables that are not always available at all weather stations. The Hargreaves equation (HG) requires only measured daily air temperature data and computed extraterrestrial radiation for ET estimates. Unless it is regionally calibrated, however, HG often tends to systematically overestimate or underestimate ET. This equation was evaluated under semiarid conditions in the Bekaa Valley of Lebanon using 16 years of complete daily climatic data from the Terbol weather station. The HG results were compared to ET estimates obtained from the FAO56 Penman Monteith equation (PM), which was used as a standard. The original HG equation overestimated ET by 23, 17, and 12% for daily, weekly, and monthly ET, respectively, as compared to PM. The results of a simple linear regression applied to obtain the calibrated HG coefficients for all three time steps showed that the calibrated equation improved the accuracy of the estimation to 3, 2, and 1% difference from ET computed by the PM method, with root mean square error (RMSE) of 0.48, 0.33, and 0.25 mmd-1 for daily, weekly, and monthly ET, respectively. Additional improvement in HG estimation accuracy was achieved by adding the wind speed using a backward variable selection method. This method resulted in only a slight improvement, reaching less than a 1% difference for all timescales and RMSE of 0.46, 0.31, and 0.22 mmd-1 for daily, weekly, and monthly ET, respectively. Thus, when only temperature data are available, the calibrated HG equation is recommended for use in the semiarid conditions of Lebanon, and when complete and reliable weather data exist, the use of the standard FAO56 PM equation is recommended. © 2013 American Society of Civil Engineers.