To improve the accuracy and efficiency of turbine rotor fault diagnosis, a new method of fault diagnosis based on the biogeography-based optimization with chaos (CS-BBO) and support vector machine (SVM) is introduced. Firstly, the chaos theory is introduced into the biogeography-based optimization algorithm (BBO), the CS-BBO algorithm is obtained. Then, the optimal parameters of the SVM diagnostic model are obtained through the CS-BBO algorithm, and optimization model enhances the learning ability and generalization ability of SVM. Finally, the validity of the optimization model is verified by the experimental data of 4 kinds of states from ZT-3 rotor test bench to simulate the turbine rotor fault. The results show that the optimized model of SVM obtained by CS-BBO algorithm can be used to diagnose the fault of the steam turbine rotor accurately and efficiently. Compared with the optimized SVM model, which is obtained by the biogeography-based optimization algorithm, the accuracy and efficiency of fault diagnosis of this method is higher.