For this work an experiment was established in the Campus of University Federal de Vicosa, with the species Carioquinha of bean crop. The experiment was divided into two plots, irrigated with frequencies of five and ten days. Crop evapotranspiration (ET) was estimated by mean of the methods of soil water balance, Class-A pan evaporation (FAO-24) and Ritchie's model. The aim of this work was to test the model proposed by Ritchie (1972) for the soil and climate conditions of Vicosa, MG - Brazil; analyzing the sensitivity in the model input parameters U and alpha (soil physical parameters), leaf area index (LAI) and maximum evapotranspiration (ETMAX); and developing a software to easy Ritchie's model using. The analysis of the results showed that the Ritchie's model properly simulated ET in the plot irrigated with high frequency and overestimated the ET value in the plot irrigated with medium frequency. For the range of values analyzed, the parameter LAI presented low sensitivity, the parameters U and alpha presented medium sensitivity, whereas the variable ETMAX presented high sensitivity.