This paper addresses the stability and tracking control problem of quadrotor unmanned flying vehicle (UAV) in the presence of uncertainty. Adaptive autonomous sliding mode tracking system is designed by combining robust and adaptive control theory. Lyapunov analysis shows that the proposed algorithms can guarantee asymptotic convergence of the tracking error of the linear and angular motion dynamics of the vehicle. Compared with other existing adaptive backstepping design, the proposed method is very simple and easy to implement on an actual system as it does not require multiple design steps without augmented signals and a priori known bound of the uncertainty. To illustrate the theoretical argument, evaluation results on custom made quadrotor UAV system are presented for real-time applications.