Accounting for the characteristics of E-commerce Website personal service and the features of users' and goods' similarities distribution, an E-commerce recommendation method based on clustering using genetic algorithm is designed. By using a composite weight matrix to integrate the situation of users purchasing, this method improves the result of clustering, and the result of clustering reflects the similarities of users and goods more accurately. This method is accorded with the reality of E-commerce Website personal service and is perfect for users' and goods' clustering computing on E-commerce Website recommendation.