The vibration reduction efficiency of a tuned mass damper （TMD） is closely related to the inherent parameters of the structure and TMD. Accurate identification of the structure and TMD inherent parameters from the response of the structure-TMD coupling system is necessary for the evaluation of the vibration reduction performance of in-service TMD. This paper offers a pa rameter identification approach based on NSGA-Ⅱ（Nondominated sorting genetic algorithm） that can identify the parameters of ‘bare structure’ and ‘bare TMD’ from the coupled structure-TMD response in order to solve the issue of state evaluation of TMDs in service. The structure-TMD coupling equation is constructed. It is reduced and transformed into a two-degree-of-freedom system of the structure-controlled mode and TMD coupling. Two objective functions are constructed by means of the system state space matrix. The genetic algorithm is used to find the optimal solution corresponding to the minimum error between the theoretical value and the experimental value， so as to identify the modal parameters of the structure and TMD. The numerical simulation anal ysis of parameter identification of single-degree-of-freedom structure-TMD coupling system and multi-degree-of-freedom structure TMD coupling system is carried out. The results show that the proposed method can accurately identify the inherent parameters of structure and TMD from the dynamic system response of the coupling system.