Abstract: Because of the high request to classifies performance of people and the implementation complexity of multiple classifiers ensemble approach，this paper proposes an new method of selective multiple classifiers ensemble which considers of the accuracy of individual classifier and diversity among individual classifiers.This algorithm first chooses the more accuracy classifies from the production base，then chooses more different ones using diversity measure before integration.The result of the UCI database experiment demonstrate that the method is better than the Bagging method，and it is very good and useful for classification.