An adaptive sliding-mode control scheme based on extended state observer (ESO) is proposed for permanent magnet synchronous motor (PMSM) chaotic system with some immesureable states. The adaptive sliding-mode control and extended state observer theory are combined in the developed controller, and thus the restriction that all the states in the PMSM should be completely measured is canceled. Through a simple coordinate transformation, the PMSM chaotic model is transformed into a Brunovsky canonical form, which is more suitable for the sliding-mode controller design. In the presence of unknown states and upper bound of nonlinear uncertainty, the ESO is employed to estimate the unknown states and the nonlinear uncertainty. Then, the adaptive sliding-mode controller is designed to ensure that the system states can converge to zero rapidly and stably. Simulation results show that the proposed controller can improve the chattering problem of the sliding-mode control and enhance the robustness of the system.