An adaptive control strategy based on the optimal sliding surface is designed for multimotor driving servo systems in this article. By dividing the system into load subsystem and motor subsystem and decomposing the tracking performance index, the optimal control problem of a high-order systems is simplified into a second-order system optimal control problem. Then based on the decomposed performance indexes, the optimal sliding surfaces are determined by solving two algebraic Riccati equations. Combined with an adaptive gain, an optimal adaptive tracking controller is designed, which can reduce the settling time and the energy consumption with a small overshoot. In addition, to obtain the unmeasurable angular velocities of motors and the load, an leaky echo state network-based observer is applied that ensures fast and accurate estimation. Furthermore, to solve the synchronization problem of the multimotor driving servo systems, the mean relative coupling synchronization controller is designed. The simulation and experimental results have demonstrated that the presented control scheme is effective.