Abstract: Reference evapotranspiration (ET₀) reflects the integrated impacts of weather conditions on the evaporation and transpiration, and serves as the basic data for crop irrigation and water allocation in irrigation areas. The estimation of the reference evapotranspiration with weather information, if feasible, will be useful to irrigation scheduling, especially in the areas where no instrumentations installed. In this study, an ET₀ prediction model was reported. The model is based on adaptive neuro-fuzzy inference system (ANFIS) and uses commonly available weather information such as sunshine condition and daily maximum temperature to forecast ET₀. The daily meteorological data from 1995 to 2003 at Daxing County, Beijing, were used to train the model, and the data in 2004 were used to predict the ET₀ in that year and to validate the model. The ET₀ in training period (Train-ET₀) and the predicted results (Test-ET₀) were compared with the ET₀ computed by Penman-Monteith method (PM-ET₀). The results indicated that the PM-ET₀ values were closely and linearly correlated with Train-ET₀ and Test-ET₀ with regression coefficient of 0.9015 and 0.8366, respectively, and showed the higher significances (α=0.01) of the Train-ET₀ and Test-ET₀. The results indicate the feasibility of using the convenient model to resolve the problems of agriculture irrigation with intelligent algorithm, and more accurate weather forecast, appropriate membership function and suitable fuzzy rules.