Abstract: Complex systems modeling method based on the combination of MTS fuzzy model with the radial basis function is presented in this paper.A direct correspondence is shown between MTS fuzzy model with the determining consequence structures and the radial basis function networks,.The premise structures of MTS model and consequence parameters can be separated in term of this direct correspondence,So we can obtain the fuzzy model using the combination of the radial basis function learning and the square recursive algorithm.The proposed approach has the advantages of simplicity,flexibility and high accuracy.Simulation studies are presented to illustrate the effectiveness of the method.