In this paper, a novel test rig for a quarter car suspension system of Samand with McPherson mechanism is fabricated and its elasto-damping elements are dynamically identified. The inputs of test rig are road roughness and its acceleration and the outputs are sprung mass acceleration, un-sprung mass acceleration, suspension deflection, and tire deflection which are recorded by sensors. The test rig of suspension system includes McPherson mechanism with nonlinear spring and damper. This system is categorized as a multi-input-multi-output (MIMO) identified system. The nonlinear least squares iterative method, as a gray-box identification method, is used for finding the elasto-damping coefficients of tire and suspension elements. In this method, a nonlinear mathematical model is considered for the system and its parameters are calculated using the test rig data. The Levenberg–Marquardt algorithm (LMA) is used to solve the non-linear least squares problem. The outputs of the identified nonlinear model are compared with the measured experimental data. As a result, the test rig outputs are followed by the outputs of the identified model with acceptable errors. The compared results indicate a good performance of the proposed model to estimate the behavior of the nonlinear suspension elements.