It is the key link to maintain the relative competitive advantage to accurately classify the clients and take precise marketing strategy in the increasing competitive commercial market of Chinese agricultural products. The RFM model was introduced in this paper to construct the classification index of e-commerce clients of agricultural products and the client clustering was inspected in combination with the RFM analysis method of SPSS software and K-Means algorithm, it was found that the conclusion was the same to adopt these two methods: the e-commerce clients of agricultural products could be classified into 8 types and corresponding marketing strategy was proposed pertinent to clients of different types.