Abstract： In this paper, the identification and modeling of coupling dynamic stiffness in joints of machine tool was discussed. The theoretical model was built by Frequency Response Function(FRF), and the identification formulas were derived based on the mechanical equilibrium condition and the displacement compatibility condition. In the process of identification, the FRF which was difficult to be measured was considered as an intermediate variable to avoid introducing equation errors. The inconsistent equation containing the identification relationship was established, and the equivalent of coupling dynamic stiffness was obtained by using generalized inverse matrix. The coupling relationship between stiffness was probed, and the method of modeling the coupling dynamic stiffness was derived, while the models containing the coupling relationship between stiffness in varying degrees were built. Finally, field tests were conducted based on the LMS system and good results were acquired, which verified the feasibility of the proposed method. By comparing the prediction results of different dynamical models, the conclusion was made that the coupling relationship could have a significant impact on estimated performance in low frequency, and that the coupling relationship should be considered in dynamics modeling.