In this paper, we have presented a method for using multiple local models to both approximate and control nonlinear systems. This method uses model weighting functions that are based on the linear distance from the nearest local models. The model regimes can be chosen such that existing models may be used, while few additional models must be identified. Both the dynamic and steady-state characteristics of the nonlinear system may be approximated within this framework. The local models are also not required to be linear in form. The multiple local model controller performs setpoint tracking and disturbance rejection.