This paper presents a modified nonlinear dynamic inversion (NDI) method for aircraft flight control system design. A hierarchical procedure is formulated to separate the control system into three subsystems according to different time scales appearing in flight dynamics. Using the proposed modification approach, no differential term is required in the formulation to bring the implicit control variables into explicit presentation. The differentiation is generally avoided in the flight control design to eliminate errors caused by sensor noise or dynamic model uncertainty. To validate the theory, in this paper, the control system design is verified using an unmanned aerial vehicle (UAV). A numerical simulation is conducted for flying in normal flight conditions as well as in severe weather conditions to demonstrate the improved performance characteristics in flight control system design. The result is acceptable to lead for real flight experiment in the next phase of UAV development.