Abstarct: Characterization and modeling of the performance of magnetorheological elastomer is the purpose of this study. Magnetorheological elastomer or MRE is a smart viscoelastic material with variable rheological properties. By use of experimental data that has done at different loading conditions such as 1, 3, 5 and 8 Hz of input frequency and 0, 100, 170, 220 and 260 mTesla of magnetic flux density and 1, 2, 4 and 8 percent of strain amplitude, the reaction of the material under different condition is studied. Removing noise from experimental data is done by use of wavelet transform at MATLAB software. The dynamic characterization of magnetorheological elastomer is experimentally explored at various input conditions of normal tests and the input condition dependencies of storage modulus and loss modulus of sample has explored. A generalized Maxwell viscoelastic model is improved to explain the relationships between normal stress and strain of magnetorheological elastomers baxsed on input frequency, strain amplitude, and magnetic flux density. The coefficients of the model under various input conditions, such as magnetic flux density, strain amplitude and input frequency, are calculated by using genetic algorithm. Finally, comparisons between the proposed model and experimental data are discussed, and the results proves the success of phenomenological model in describing the behavior of magnetorheological elastomers under tensile-compressive loadings, very well. The presented model is useful to simulate the performances of magnetorheological elastomer baxse devices under normal small strain loadings.