This paper investigates the problem of model reduction for interval type-2 (IT2) fuzzy systems subject to D stability constraints. The membership functions and the number of rules can be freely chosen and they are different between the original system and the reduced-order system. By introducing some slack matrices and utilizing Lyapunov stability theory, the existence condition of model reduction is obtained to guarantee that the reduced-order model can approximate the original system with an Ho, performance. The parameters of the reduced-order system in the condition can be obtained by standard software. Finally, some simulation results are provided to demonstrate the effectiveness of the proposed results. (C) 2015 Elsevier Ltd. All rights reserved.