The purpose of this study was to identify a mechanomyogram (MMG) system with subspace-based state space model identification (4SID), a black-box system identification method. The input data consisted of the electrical stimulation of the common peroneal nerve, which made the anterior tibial muscle contract. The output data consisted of the evoked MMG. The system was well identified with the sixthto tenth-order models. The transfer function was factorized to the second-order model describing the mechanical systems consisting of mass, viscosity and elasticity. The coefficients of the second-order model were evaluated, and the coefficients reflecting the elastic properties were classified into two groups: The group reflecting active muscle elements increased as the contraction level increased, while that reflecting passive elements such as skin or fat did not. These results suggest that the mechanical properties of both active muscle and passive tissue can be evaluated by the proposed method.