This paper assesses the effects of (1) the gravity load‐resisting system (GLRS) modeling approach, (2) the seismic force‐resisting system (SFRS) modeling approach, and (3) the uncertainty of the model parameters of the constitutive law of the longitudinal reinforcing steel of the SFRS on the seismic responses of a 12‐story reinforced concrete wall building with force‐limiting connections. This is achieved by conducting nonlinear numerical earthquake simulations. The seismic responses of the building models with force‐limiting connections using two GLRS modeling approaches, (1) a moment frame system and (2) a pin‐base lean‐on‐column system, are compared. The seismic responses of the building models with conventional connections and force‐limiting connections, respectively, using two SFRS modeling approaches, (1) a distributed‐plasticity modeling approach and (2) a lumped‐plasticity modeling approach, are compared. A joint probability density function for the ASTM‐A615 Grade 60 steel available in the literature is used to conduct an uncertainty propagation analysis through Monte Carlo simulation. The uncertainty in the steel model parameters is propagated to the seismic responses of the building models with conventional connections and force‐limiting connections, respectively. The distributions of the mean values of the peak structural responses of the building models are studied. The effects of the GLRS modeling approach on the seismic responses are not significant in the context of seismic performance‐based design and assessment of buildings with force‐limiting connections. The effects of the SFRS modeling approach and the uncertainty in the steel model parameters on the floor total acceleration and force responses are reduced by including force‐limiting connections.