A new finite-time control method based on a sliding mode for a multirotor unmanned aerial vehicle (UAV) is developed to improve both the transient and steady-state responses, including overshoot and steady-state error in the presence of uncertainties and external disturbances. First, a virtual control with nonlinear sliding manifolds is designed to achieve position-tracking capability, as well as to guarantee the fast convergence of the UAV to a desired position. Furthermore, an ultimate control is developed for the desired attitude-tracking performance. Various uncertainties, including torque due to the discordance between the centre of mass and rotation and wind disturbances are considered. The Lyapunov stability theorem is then applied step-by-step to prove the asymptotically stable and finite-time convergence in position and attitude controllers. Second, the proposed controller is implemented in an open-source hardware platform for a quadrotor UAV. Both numerical and experimental results are compared to validate the tracking performance for attitude and position control, as well as robustness under disturbances. INDEX TERMS Finite-time control, sliding-mode control (SMC), unmanned aerial vehicle (UAV).