Abstract: This paper gives a brief introduction of the generalprinciple of model updating as well as the methods of model reduction andmodal expansion closely related to model updating. Some representativereferences are singled out and introduced with particular attentionto three kinds of primary methods of modelupdating, namely, direct method and parametric method, which are known astraditional methods, and updating method with neural networks. Emphases areput on their merits in application and corresponding disadvantages.Finally, some recent unresolved problemsare discussed.