In general, the performance of movable nozzle used for thrust vector control in solid rocket motor is estimated on the basis of the effective pivot of nozzle. However, it is nearly impossible to define the exact effective pivot by the mathematical model or experimental technique owing to pivot dynamics. In this paper, pivot dynamic properties were investigated by ADAMS simulation technique and trajectory of the exact effective pivot was modelled by the artificial neural network. Comparison of the proposed method was made with the virtual movable nozzle (computer simulation) to verify the method, and showed good agreement. Therefore, the proposed method will be applicable to predict the effective pivot of movable nozzle during bench or ground test.