Performance-based earthquake engineering requires numerical models that are able to reliably predict the expected component behavior of buildings under seismic excitations. The nonlinear modeling guidelines provided in ASCE-41-13 have historically been used in this context by practicing engineers. However, for steel wide-flange columns, the recommendations are primarily based on experimental data that was fairly limited until recently. This paper addresses many of the limitations present in ASCE-41-13 for the modeling of steel wide-flange columns. Recent experimental data from large- and full-scale tests on wide-flange steel columns, supplemented with detailed finite element studies, has been collected for a wide range of column cross-sections and applied compressive axial loads. Multiple regression analysis is utilized to develop multivariate predictive equations for the seismic response of steel wide-flange columns. It is found that the primary contributors that dominate the column response are the local web slenderness, member slenderness and the applied axial load ratio due to gravity loading. Guidelines for developing nonlinear component models for steel columns that can be directly used in both nonlinear static and dynamic analysis procedures are provided. The predicted responses are validated against test data that was not part of the original database. Furthermore, recommendations are proposed for modeling wide-flange columns subjected to varying axial loads as well as bidirectional loading; and a new limit for force-controlled elements is proposed for columns under high axial compressive loading and lateral drift demands.