In this paper, the disturbance observer-based controller (DOBC) design problem is investigated for a class of uncertain nonlinear parameter-varying (NPV) systems subject to unknown uncertainty and unmeasurable state variables. First, aiming to deal with the unknown uncertainty, an efficient scheme for treating the uncertainty as an unknown disturbance is given. Second, based on the transformed model with unknown disturbance, a novel observer is presented to estimate the unmeasurable state variables and unknown disturbance, which is further used to design the state-and-parameter-dependent (SAPD) controller. By using the correspond closed-loop systems and Lyapunov stability theory, some SAPD conditions on designing the observer and controller are established by linear matrix inequalities (LMIs) forms. Based on sum-of-squares (SOS) techniques, these LMIs can be effective solved Remarkably, the SAPD controller and the SAPD observer are designed independently, which can significantly reduce the complexity of the disturbance-based control algorithm. Finally, some simulations and comparative results in two examples are provided to prove the feasibility of the proposed approach.