Abstract: A surrogate-based inverse procedure is presented for characterizing the material property of functionally graded material（FGM）. A forward model was built by the finite element software and the dynamic displacement responses excited by the load was obtained. A response surface model was established based on the responses of the sample points selected by employing design of experiment， the substitute for the finite element software as the forward operator and the genetic algorithm was employed as the inverse operator. The maximums of the displacement responses with different levels of noise were used as the inputs for the whole procedure，and the material volume fraction parameters were finally determined. The dynamic displacement responses excited by the single sin load in an actual SiC C functionally graded beam were applied to determine the volume fractions of the SiC-C beam through the genetic algorithm. The numeric example demonstrates the efficiency of the presented method. The surrogate-based inverse method improves the calculating efficiency by avoiding repetitious forward calculations.