This paper proposes a multiple model approach for a nonlinear multi-input, multi-output (MIMO) system identification (SI) of smart structures equipped with magnetorheological (MR) dampers. The proposed model is developed through integration of MIMO autoregressive exogenous (ARX) input models, Takagi-Sugeno (TS) fuzzy model, weighted linear least squares estimators, and data clustering algorithms. Nonlinear behavior of the structure-MR damper systems is represented by a set of linear MIMO ARX input models whose operating regions are blended by TS fuzzy sets. To demonstrate the effectiveness of the proposed MIMO ARX-TS fuzzy model, a 20-story high-rise building employing MR dampers is investigated. It is demonstrated that the proposed approach is effective in modeling nonlinear behavior of the structure-MR damper system subjected to a variety of disturbances. Comparison with high fidelity data proves the viability of the proposed approach in control engineering setting.