In this paper, a direct adaptive self-structuring fuzzy control (DASFC) scheme is proposed for tracking a class of nonlinear uncertain systems. In the DASFC, the ideal control law with unknown dynamics and uncertainties is approximated by a self-structuring interpretable fuzzy system (SIFS) which can automatically self-construct transparent fuzzy rules together with interpretable fuzzy sets projected in each dimension. Moreover, adaptive approximation error compensation and projection-based adaptive laws for all free parameters not only to address the robustness issue due to the SIFS-based approximation error, but also to guarantee globally asymptotical stability of the overall DASFC control system.