The Revised Universal Soil Loss Equation (RUSLE) is used for predicting soil erosion. Rainfall erosivity (EI) in this equation is related to storm type, amount and intensity so it should be determined from rainfall characteristics. In the present research, data from 180 recording rain gauge stations throughout the I.R. of Iran were analyzed and single storm, daily, monthly and annual erosion indices were calculated and estimated by different simple models. For the single storm erosion index, the models EI=a Peb/Dg and EI=pPe q were modified. Coefficients of these simple models were found to be elevation, longitude and latitude dependent. Therefore, multiple regression equations were used to estimate these coefficients based on the elevation, longitude and latitude of the stations. For the daily erosion index, a power function based on daily rainfall is presented. The values of coefficients for this equation were dependent on the elevation, longitude and latitude of stations and they are estimated by the given multiple regressions. For the monthly erosion index, a simple model based on the monthly maximum daily rainfall was proposed. The values of the coefficients for this equation were also determined by the given multiple regression equations. The coefficients of the Arnoldus model was modified for the study region to estimate the annual erosion index using monthly and annual rainfalls. The coefficients of this model are elevation, longitude and latitude dependent and are estimated by multiple regression equations. According to the simple model, for the monthly erosivity estimation with modified coefficients the annual iso-erosivity map was drawn for the study region. The range of annual erosivity for the study region was similar to those reported for a neighbor country (i.e., Iraq).