Vibrations are experienced as unwanted dynamic motions and the industry takes a lot of care to reduce these motions to improve comfort and to increase vehicle component durability. The properties of rubber, being generally soft while showing large inherent damping, makes the material a perfect choice to use in vibration isolators. To minimize vibrations the transfer path from vibrational source to the driver almost always includes a highly damped material like rubber. With the current trend of electrification in the automotive industry, higher order vibrations are becoming more pronounced. Although the produced sound is not of a high octave with respect to traditional combustion engines, they are experienced as unpleasant. Considering a large frequency range, the dynamic characterization of a rubber object is proven to be a difficult topic. In the search of improved rubber characterization new methods arise. The goal of this research is to provide a new approach for extracting material properties from the experimental impact method. Literature research is done on the rubber specific behaviour and how this affects the modelling approach. In the research itself both numerical and experimental characterization is used. Through impact testing the receptance of the rubber object is measured and by applying inverse substructuring the dynamic stiffness is obtained. The Finite Element model is used to directly obtain the dynamic stiffness. The numerical and experimental results are made comparable by applying the virtual point transformation method. With the finite element material model being the optimization variable, the numerical dynamic stiffness is optimized to match the experimental results. The rubber properties are obtained from the optimized material model. The presented inverse approach makes it possible to use the impact method for characterizing rubber material properties. It therefore broadens the abilities of impact testing used for rubber characterization. With the new method the drivingpoint properties can be predicted which do not depend on complex decoupling methods. The obtained material properties are material, but not geometry related. The findings can be used for different geometries of the same rubber. The new approach has the ability to significantly reduce experimental effort because the material properties can be extracted from one single loadcase, whereas the experimental should excite all loadcases to fully characterize the dynamic stiffness.