This research study is aimed at evaluating the accuracy of the displacement coefficient method (DCM) of FEMA 440 and associated nonlinear static procedure (NLSP) for actual buildings with soft story mechanism and various ductility capacities. The DCM and associated NLSP are evaluated using two existing seismically vulnerable buildings with soft story mechanism. The buildings are first retrofitted using a ductile steel-brace-link system to represent those with good ductility capacity and then retrofitted with RC squat infill shear panels (SISPs) to represent those with relatively poor ductility capacity. The evaluation of the DCM of FEMA 440 and associated NLSP is then performed by comparing the roof displacements (target displacements), maximum interstory drifts, and maximum plastic hinge rotations of the original and retrofitted buildings obtained from NLSP (at the target displacement level of DCM) with those obtained from nonlinear response history (NRH) analyses for three different seismic performance levels. It is observed that the DCM, and hence, the NLSP fail to accurately predict the NRH analyses results mainly due to uncertainties in the coefficient C1 of the DCM in the short period range, the inability of the DCM to capture the failure of structural members beyond a certain lateral displacement or plastic rotation limit and associated soft story mechanism. Copyright © 2013 John Wiley & Sons, Ltd.