The process of erosion, traort and deposition is of major problems and difficulties in the development of water resources and watershed plans in our country. Knowing about the soil erosion and sedimentation in rivers in the catchment management is of great importance in all managerial plans. Plan that have been implemented without regard to this issue did not have the needed efficiency and caused a lot of damage to the country's capitals. So, the amount of sedimentation of rivers with regard to its damages are of main requirements in the field of basin management. Not having enough sediment-Measurement stations along with new studies and limitation of years with statistics are three major reasons to prevent from achieving reliable statistics in this field. Therefore, calculating the rate of sediment in basins without station, requires the use of regional analysis methods to achieve appropriate models for estimating sediment load. Regional analysis means the use of collected data from specific locations and extending it to the entire surface of a region. Therefore, this research aimed to analyze the regional modeling of sediment in the North Karun basin. In this study, 12 hydrometric and Sediment Measurement stations with sufficient statistics and good distribution in the region were selected and the amount of suspended sediment have been calculated through two ways of Sediment rating curve and combining the monthly, seasonally and yearly discharge statistics. Besides, in the upstream basins of mentioned stations, 35 physiographic, soil, land use and climate features have been calculated as factors of production and multivariate regression analysis have been accomplished using two approaches. 1- With all variants 2- with all selected factors. And the most suitable statistical relation between suspended sediment and basin features at 33 temporal models along with the error rate of each proposed model have been obtained using coefficients of AME, MAE, RMSE, R2 and NS. The results of this study showed that predictive models of suspended sediment acquired from regression method based on all variables gives better results with less error coefficients. Besides, the models which their observed sediments have been completed based on the annual sediment rating curve show a higher coefficient of determination. Keywords : Factor analysis, Sediment data, Multivariated regression analysis, Error coefficients, Modeling