Precipitation is one of the most important aerology parameters whose forecasting has significance in the prediction of drought and management of catchment. Today by creating evolutionary algorithms, this is done quickly and more accurately. One of these algorithms is the Imperialist Competitive Algorithm. In this study, using data for 15 years (1996-2010) in three synoptic stations in Kerman, the performance of this algorithm in forecasting daily precipitation is checked. The results show relatively good performance for this algorithm in precipitation, with a correlation coefficient of 0.8605, root mean square error of 1.029 and Willmott index of agreement of 0.96. However, because of the large number of dry days, accumulation of points are located in the origin of the coordinates.