This study describes and demonstrates the modelfollowing control using a robust gain matrix based on the normalized plant. The normalized plant is constructed by inserting the matrices for normalizing into the original plant, rendering it independent of the plant parameters. The robust gain matrix also adopts a switching function determined by the linear quadratic regulator (LQR) method. The robust gain matrix is then incorporated into a model-following control and evaluated in simulations of the nominal plant, the plant with a modeling error, and the plant with disturbances. The superior performance of the proposed method is demonstrated in loaded and unloaded DC motor experiments.