The wheel and rail receptances derived in the preceding companion papers [1-3] are combined in a model of wheel-rail interaction. For this model the receptances of the contact region (local deformations, etc.) are also required, and these are derived first for six degrees of freedom. These include the effects of the (linearized) Hertzian contact spring, and Kalker's creep force formulation including frequency-dependent creep coefficients. Results of the interaction model are compared with experimental data taken from a series of train running tests, and appear to give reasonable agreement, particularly for the vertical responses. However, some discrepancies remain, particularly for the lateral responses of the wheel and rail at the contact. These appear to be the result of inadequacies in the lateral-vertical cross-receptances of the wheel and rail. Alternative forcing models are also considered (relative force, and absolute force/displacement inputs), but appear to be inferior to the relative displacement model, which has been used throughout the rest of the study. The lateral movement of the contact patch has also been considered, and has been found to be potentially significant as an additional source of excitation.