Abstract: Fabric’s shearing property has a close relation with the factors such as fiber，yarn and fabric structures.In order to design the fabric’s shearing property scientifically，a new method is proposed to predict the fabric’s shearing property with support vector machine.The nonlinear relationship between fabric structural parameter and fabric’s shearing property is presented，and the relative model is built.After normalized，the sampling data are inputted into the model.Then cross-validation is used to select the optimal parameters and the obtained optimal parameters are used to predict the fabric’s shearing property.Compared with BP Artificial Neural Network（ANN） method，the predicted results show the presented method is more accuracy.