With the advent of performance-based earthquake engineering (PBEE), the need for reliable prediction of earthquake-induced collapse of structures is essential. Despite the significant progress that has been made towards this goal, there are several hurdles that still need to be overcome. Deterioration models, which are currently used for simulating structural collapse, are largely based on available test data that featured subassemblies with overly simplified boundary conditions. These experiments did not ac-count for the force redistribution occurring within structural systems after the onset of nonlinear geo-metric instabilities under cyclic loading. For the case of composite steel moment resisting frames (MRFs), which is the primary focus of this thesis, the influence of the axial restraint provided by the slab continuity and framing action on the seismic behavior of composite steel MRFs has been recognized in prior work. Nevertheless, these effects have neither been thoroughly investigated nor quantified at lateral drift demands associated with dynamic instability of composite steel MRFs under seismic loading. In this doctoral thesis, a unique experimental program of a two-bay composite steel MRF subsystem was conducted at full-scale. The experimental program, which was corroborated by continuum finite element analyses, aimed at comprehending the role of the underlying physical mechanisms, associated with the slab continuity and framing action, on the hysteretic behavior of composite steel MRFs with a particular emphasis at large deformations associated with collapse. The research findings suggest that the slab continuity limits the extent of local buckling and beam shortening within the dissipative zones of composite steel MRFs. While nonlinear geometric instabilities attributable to local buckling and concrete crushing are pronounced in composite steel beams at large inelastic deformations, the framing action and slab continuity preserve the overall stability of the structural system even in cases that ductile cracks form within a dissipative zone. The test results suggest that the presence of the transverse beams within the floor system should be factored in capacity design principles of composite steel MRFs. The experimental findings also suggest that methods for computing the effective slab width in composite steel MRFs should be reassessed. The available data along with information from prior experiments informed the development of practice-oriented engineering models for the seismic assessment of composite steel MRFs. Moreover, a macro-model was proposed for simulating the hysteretic behavior of composite steel beams under cyclic loading. The model, which was validated thoroughly with available experimental data, simulates explicitly the cyclic deterioration in strength and stiffness of composite steel beams and their respective beam-slab connections. The proposed macro-model was employed to benchmark the seismic col-lapse risk of prototype buildings with composite steel MRFs, as well as the repairability of beam-slab connections within the framework of PBEE.