A 4-story steel moment-frame building designed according to ASCE 7 was used in a numerical parameter study to assess the effects of modeling features on peak drift demands. Features studied included strength, stiffness, ductility, and degradation along with several hysteretic models. Attention was given to ASCE 41-type backbone curves. Of particular interest was exploring the effects of degradation, in which an adaptive backbone curve was used to capture both in-cycle and cyclic degradations. Incremental dynamic analyses (IDAs) were performed using a suite of earthquake records to assess the response over a range of shaking intensities. It was found that in-cycle degradation had more influence on the response compared to cyclic degradation for the set of ground motion records that were employed. Moreover, use of the monotonic backbone alone, with its in-cycle degradation, was sufficient. In addition, it was found that increasing strength, stiffness, and/or ductility resulted in decreased peak drift demands, whereas modifying the hysteretic type (elasto-plastic, stiffness-degrading, and pinching) had little effect on peak drifts. These findings indicate that using backbone curves based on envelopes of first-cycle test data, as done in ASCE 41, can result in overly conservative seismic response predictions.