Prediction of regional logistics requirement provides a basis for the plan of regional logistics. In the study, support vector regression is presented to predict regional logistics requirement. The regional logistics data from 1996 to 2006 in Shanghai municipality are used as the application data of support vector regression. The comparison of prediction error between BP neural network and support vector regression is applied to testify the superiority of support vector regression. The comparison results show that the prediction results of support vector regression are better than those of BP neural network.