Magnetorheological elastomers (MREs) are a novel class of magneto-active materials comprised of an elastomeric matrix impregnated by micron-sized ferromagnetic particles, which exhibit adjustable mechanical properties such as stiffness and damping coefficient in a reversible manner under the application of an external magnetic field. MREs are solid state of magnetorheological (MR) materials. In contrast to MR fluids, which provide field-dependent apparent viscosity, MREs, being a smart viscoelastic material, are capable of providing controlled field dependent moduli. Yet having a solid grasp of highly complex behavior of this active composite is a fundamental necessity to design any adaptive structure based on the MRE. This study is concerned with investigation of the static and dynamic behavior of the magnetorheological elastomers. To this end, six different types of MREs with varying contents of the rubber matrix as well as ferromagnetic particles are fabricated and characterized statically in the shear mode as a function of the magnetic field intensity. The MRE containing the highest percentage of iron particles (40% volume fraction) exhibited a notable relative MR effect of 555% with 181.54 KPa increase in the MRE shear modulus. This particular MRE was then chosen for subsequent dynamic characterization. The dynamic responses of magnetorheological elastomers revealed strong dependence on the strain and strain rate as well as the applied magnetic field intensity. Dynamic characterization is performed in shear mode under harmonic excitations under the broad ranges of shear strain amplitude (2.5-20%), frequency (0.1-50 Hz) and magnetic field intensity (0-450 mT). The strain softening, strain stiffening, strain rate stiffening and the magnetic field stiffening phenomena are identified as the nonlinear properties of MRE stress-strain hysteresis loops. Subsequently, an operator-based Prandtl-Ishlinskii (PI) phenomenological model is developed to predict the nonlinear hysteresis behavior of the MREs as functions of strain, strain rate and field intensity. The stop-operator-based classical PI model using only 10 hysteresis operators provided very accurate predictions, and it involved identification of only four parameters, which were dependent on the loading conditions. The validity of the developed Classical Prandtl-Ishlinskii model is assessed using the laboratory-measured data for MRE over a wide range of inputs. The proposed model is further generalized to predict the dynamic behavior of MRE independent of the loading conditions, which could be beneficial for controlling the MRE-based adaptive devices in real time. The results demonstrated that the proposed generalized model could accurately characterize the nonlinear hysteresis properties of MRE under a wide range of loading conditions and applied magnetic fields.