Storage Function Method is a method applied in discharge calculation with consideration on catchments characteristics. This method is commonly used in Indonesia by the Japanese experts. The discharge calculation provided by the method is highly influenced by the catchments rainfall, thus the difference of rainfall station number will effect the accuracy of the calculation result. The objective of this study is to investigate the effect of difference in the number of rainfall stations on the value of model parameters as well as the accuracy of discharge calculation using the storage function method on Kali Madiun Basin. Basically, the concept used in the storage function method is a water balance, which is analyzed using the mathematical model. The data used in this research are the 10 days hourly rainfall during the flood, and the catchment characteristics constant as the model parameters. Using the trial procedure, the catchments characteristics constant as the input for each of the numbers of rainfall station (2;3;4;5;6; and 7 stations) will provide the 10 days hourly discharge, then the result will be calibrated to the observed data. Based on the calibration result, the smallest deviation of each rainfall station will be selected in order to identify the model parameter and to provide deviation of the observed data. The investigated deviation is the different value between the calculated and the observed runoff, includes volume deviation as well as the 10 days hourly discharge deviation, which are located at the control node of AWLR Sekayu and AWLR A. Yani. Based on the calibration results, it is shown that the largest volume deviation at AWLR Sekayu is 31.64% that occurs on 2 rainfall stations, while the smallest occurs on 6 rainfall stations with deviation of 17.43%. It is also shown that the largest discharge is 37.29% and occurs on 2 rainfall stations, while the largest and the smallest discharge deviations respectively are 36.7% that occurs on 2 rainfall stations and 26.39% that occurs on 7 rainfall stations. Based on the results, it is shown that although the largest number of rainfall station is not the most accurate, but in general, as shown in the graphical results, it is indicated that more numbers of rainfall stations tend to provide better accuracy.