Abstract: This paper proposes an improved method called CLCC（Classification using Local Clustering with Cost-sensitive）which combines cost-sensitive and majority class decomposition.Using local cluster in the majority class can make a balance between the majority class and the minority class.And the introduction of cost-sensitive to the iterative learning process makes large improvements of the misclassification ratio of the minority class and dataset.Thus，the accuracy of the minority class is enhanced while the classification cost is decreased.The experimental results confirm that this algorithm is superior to the traditional algorithms as for dealing with the imbalanced problem.