Abstract: Soil erodibility K value could provide an important basis for the prediction of local soil loss and for benefit evaluation of soil conservation measures. Five methods were used to estimate the soil erodibility K values of different cropping patterns in purple hilly area. The paper aims to select appropriate estimation methods for calculating K values,which could reflect the process of soil formation and erosion characteristics in the purple soil.Results showed that:1) for the same soil parent material and land use types,the soil erodibility K values got by the five estimation methods could be lined in the order of KEPICKmodifiednomo K nomoKShiraziKTorri. The inherent differences among these K values were resulted from choosing varied soil physical and chemical properties as basic parameters for each estimation method. Soil erodibility K value of alfalfa was the smallest, indicating that alfalfa and other legumes were suitable plants for intercropping and alley cropping in local agricultural system on hillsides since they can effectively lower soil erosion sensitivity in this area. 2）For the same soil parent material and soil types, different land use types responded differently to the stability of estimation methods for K value, and the sensitivity order was alfalfa landwheat landmulberry land.Concerning the local cropping patterns with regular sloping land farming,there was no distinct differences in the stability of estimation methods.3）In the purple hilly area,the estimated K values by nomogragh model and EPIC model were close most to the standard value,and also the two methods were sensitive to the changes in soil physical and chemical properties of this area.Therefore, nomogragh model and EPIC model can be used to estimate soil erodibility K value when predicting soil loss amount and assessing soil erosion sensitivity in the purple hilly area.