An integrated adaptive-robust control methodology based on the sliding mode is developed for the control of nonlinear uncertain systems. The new approach possesses the advantages of both adaptive control and robust central, and eliminates the disadvantages of each. It has the ability to learn like an adaptive controller and the ability to handle various uncertainties and reject disturbances like a robust controller. When compared with pure adaptive control, the approach used improves the transient performance, and when compared with pure robust control, it reduces the conservatism of the design. Furthermore, it guarantees the global asymptotic stability of the closed-loop system. The paper presents the methodology and illustrates the approach using a simple 1-DOF robot in the vertical plane under gravity.