The performance of any structure during ground motion depends on the inherent material and structural properties. In the present work, parametric state space identification (N4SID) has been applied to obtain the modal parameters of three different laboratory scaled models. Models with and without infill walls have been considered for the study, where soft storey condition is simulated by placing infill walls in the first floor alone. The laboratory models have been excited on a shake table with different recorded ground motions and model floor acceleration responses have been obtained. Modal properties have been extracted using system identification technique for both bare frame model and model with infill walls. Very consistent modal parameters have been found from all the models from the experiment using the identification technique. Numerical models with similar structural properties have also been developed in SAP-2000 for bare frame model and models with diagonal strut on the first floor based on different recommendation to simulate the effect of infill walls. The recommended stiffness values for diagonal strut are quite different corresponding to different recommendation and hence the modal analyses of these numerical models with infill walls also show variable results. Therefore, an attempt has been made to follow a reliable methodology for more precise assessment of infill walls contribution, which becomes very critical for ascertaining the extent of weakness in a soft storeyed RC building.