Customer credit scoring is an important concern for numerous domestic and global industries. It is difficult to achieve satisfactory performance by traditional models constructed on the assumption that the training and test data are subject to the same distribution, because the customers usually come from different districts and may be subject to different distributions in reality. This study combines ensemble learning with transfer learning, and proposes a clustering and selection based transfer ensemble (CSTS) model to transfer the instances from related source domains to target domain for assisting in modeling. The experimental results in two customer credit scoring datasets show that CSTE model outperforms two traditional credit scoring models, as well as three existing transfer learning models.