It has been recognized that an isolated deck develops horizontal displacements of considerable amplitude during a strong earthquake. In this case the possibility of mobilizing the abutments in moderating such large amplitude horizontal response is beneficial for the safety of the structure. Thus, apart from lowering the seismic forces by the low-stiffness isolator units, the interaction between the deck and the abutments in the form of pounding for large horizontal deck response amplitudes aims at limiting through this mechanism excessive horizontal deck displacements. Such a problem was examined at the laboratory of Strength of Materials and Structures of Aristotle University using a small-scale physical representation that retains in a qualitative way the following important features: 1. A relatively stiff steel platform, representing the bridge deck, which is supported on a shaking table by two flexible supports, representing the isolator units; it is subjected to simulated horizontal earthquake motions developing large amplitude horizontal displacement response. 2. The possibility of bridge deck pounding on the abutment was introduced through a connector device that became active after the deck response exceeded a certain amplitude, introducing an initial gap within this connector. Despite the fact that these two basic response mechanisms, flexibility of isolator units and connector force-displacement characteristics, are crude small-scale representations of the actual mechanisms that are mobilized in a prototype bridge deck, the qualitative characteristics of this problems are retained. A number of simulated earthquake tests provided the necessary measured acceleration and displacement response of the model steel platform of the small-scale model and the force-displacement response of the connector and the flexible supports of the steel platform with the shaking table. This was next utilized to validate numerical simulations of this small-scale experimental representation of the bridge-deck pounding problem. By comparing the numerical predictions with the measured response of this small-scale experimental representation of the bridge-deck pounding problem it can be concluded that such numerical simulations can yield quite accurate predictions provided that the force-displacement characteristics of the isolator units as well as the force-displacement characteristics of the mechanism representing the bridge deck-abutment pounding are defined with reasonable accuracy for the prototype bridge.