In this paper an improved approach is presented for integrating backstepping control of outdoor quadrotor UAVs. The controller uses the approximated nonlinear dynamic model, while for simulation or test purposes the quadrotor can be modeled either with the precise or the simplified model. A hierarchical integrating backstepping control algorithm was constructed that has the capability of handling every effect in the dynamic model and in the meantime successfully ignores the realistic measurement noises. The hierarchical control structure consists of position, attitude and rotor control, extended with path design with continuous acceleration and/or continuous jerk. The state estimation is based on sensor fusion. Control parameters can be easily tuned. Adaptive laws are elaborated for mass and vertical disturbance force estimation. The tracking algorithm is able to follow the prescribed path with small error. The sensory system and the state estimation are prepared for outdoor applications. The embedded control system contains a HIL extension to test the control algorithms before the first flight under real time conditions.