jats:titleAbstract</jats:title> jats:pStatistical Energy Analysis (SEA) techniques have been largely employed to model vibroacoustic systems as it simplifies the equations of motion of the system optimising the computing time, a useful feature to analyse random ensembles. However, the approach is limited to linear systems. Modelling vibro-acoustic systems with nonlinear characteristics is a challenging problem as there rarely exist analytical solutions for the dynamic response. An example of a system that includes nonlinear features in the transmission path is the suspension system of a vehicle, where vibrations that are result of the interaction between the wheels and the road ultimately arise the sound pressure levels in the car cabin, which are affected by such nonlinearities. A nonlinear interface between the excitation point and a statistical structure has been included in an experimental setup that represents the suspension system with the aim of exploring key features, or otherwise, of the effect that a transmission path with nonlinear stiffness has on the structural response of a randomised statistical system to a random input. A numerical model has been developed to simulate the dynamic response of the structure by adopting the infinite plate assumption to model the statistical structure as an SEA dissipative mechanism, hence the equations of motion are largely simplified as they result in a single-degree-of-freedom second order differential equation. Numerical simulations of the model here developed agree remarkably well with the experimental data, where the averaged dynamic response and the generation of high-order harmonics are accurately estimated. Additionally, the measured loss of coherence at the frequencies where harmonics are present is also predicted by the model.</jats:p>