As for inverse problem, there exist lots of traditional mathematic and physical methods. In recent years, the burgeoning evolutionary algorithms have been widely utilized in many areas. Their global optimization, parallelism, and effectiveness are widely applied to functional optimization problems. A method based on Gene Expression Programming is presented in this paper for identifying the parameter in inverse problems. The results show that the new method is very effective for solving the inverse problem even for the case with large noise in the observation data.