Abstract | The choice of available finite element modeling conventions of reinforced concrete (RC) buildings may influence significantly the predicted inelastic seismic performance during the evaluation of existing RC frame redesigns. The problem is investigated by analyzing alternative models of a typical existing five-story RC frame which has been designed for moderate seismicity, using past seismic provisions. Different inelastic finite element models of the structure are established, taking into account a range of practical and more detailed finite element idealizations that range from the widely adopted concentrated plasticity elements to the more complex distributed damage fiber elements, accounting or not for large deformations, joint flexibility, bond or shear deterioration. Following static inelastic analysis, key performance global and local response indices (such as interstory drift and plastic rotations) which are adopted in Performance Based Design, are estimated, to quantify the uncertainty introduced strictly by the modeling conventions alone. It is concluded that key response indicators (collapse mechanism, capacity curve and maximum plastic rotations) can be grossly underestimated strictly due to the model alone, something that should be considered in the normative process by tight modeling guidelines and adequate safety factors.