Considering the adaptive tracking control for nonlinear system with partially unavailable states. Firstly, a new nonlinear system mathematical model is proposed, which is consisted of available states subsystem and unavailable states subsystem. The two subsystems are coupled with each other. Secondly, the unknown nonlinear functions and uncertain parameter are estimated. The estimated information of unknown variables is obtained. Finally, the adaptive controller only is utilized by the available states and other known bounded information. By the Lyapunov stability theory, it is proved that available states and unavailable states can asymptotically track reference goals respectively, with the operation of adaptive controller synthesized. A simulation example is given to verify the effectiveness of control method proposed. Compared with the controller designed based on the state observer, the controller proposed by using only available states has the better control preference without introducing the state observer to estimate unavailable states.