Nowadays,with the increasing competition between banks,it is very important to improve the bank's competitiveness by effectively excavating potential clients and providing differentiated services.The decision tree algorithm is proposed to data mine the possible 21 important attributes which affect bank clients' long-term deposit subscription.A Portuguese retail bank is addressed,with data collected from May,2008 to November,2010.The decision tree model is constructed to reflect the important factors in a banking client deposit subscription business.Results show that the significant factors which affect client's long-term deposit subscription are target number of bank agents,marketing duration and month.Such knowledge greatly reduces the marketing range of potential clients for term deposit thus improves the bank efficiency.