When dynamic performance of structures is assessed experimentally, most experimental studies consider variation in intensity by increasing the record of peak ground acceleration or displacement at different values. However, severity of earthquake motion is not only a function of the concerns instrumental intensity, but also the frequency content, duration and the amount of energy imposed on the structure. To fulfill dynamic test conditions, this paper describes and discusses a novel method for selecting several seismic events characterising the seismic area and fulfilling particular similitude requirements. The method included using an earthquake record as an empirical Green's function for simulating larger magnitude events. Three seismic events used for shake table testing of reinforced concrete walls for low-rise housing are presented and discussed to validate the proposed method's results.