An adaptive dynamic inversion control formulation is presented that takes advantage of the inherent dynamic structure of the state-space description of a large class of systems. The formulations impose the exact kinematic differential equations, thereby restricting the adaptation process that compensates for model errors to the acceleration level. The utility of this formulation is demonstrated for the problem of fault tolerance to actuator failures on redundantly actuated systems. The approach incorporates an actuator failure model in the controller formulation, so that actuator failure can be identified as a change in the parameters of the failure model. Tracking of reference trajectories is imposed, and initial error conditions and structured parametric uncertainties are incorporated explicitly in both the plant parameters and the control influence matrix. A numerical example consisting of a nonlinear model of an F-16 type aircraft with thrust vectoring is presented. Simulation results show that the fault-tolerant adaptive controller is capable of simultaneously handling parametric uncertainties, large initial condition errors, and actuator failures while maintaining adequate tracking performance. N recent years, there has been much interest in the development of reconfigurable control systems that can accommodate actuator failures without compromising mission integrity. There has been substantial progress in the development of real-time failure detection and isolation algorithms, system identification after failure, and control reconfiguration techniques in aerospace applications. In Ref. 1, a survey of various reconfigurable flight control methodologies is presented and it is shown that most traditional reconfiguration flight control approaches rely on failure detection and isolation. The complexity of such a system with this feature grows with the increase in the number of failures, and there tends to be a significant possibility of false alarms. 1,2 A different approach to reconfigurable flight control is based on adaptive control theory, in which the adaptive control structure implicitly reconfigures the control law using adaptive estimates of the altered dynamics after failure. 3 In Ref. 3, an adaptive control scheme is presented that uses a linear approximation of the plant model to compute the control, and a neural network based adaptive control law for flight reconfiguration has been developed and successfully flight tested. 4−6 A robust fault-tolerant controller has also been developed to reject state-dependent disturbances. 7 The approach presented in this paper uses a structured nonlinear adaptive dynamic inversion control methodology. Instead of using an explicit failure detection and isolation algorithm, this methodology is based on the adaptive control theory where the controller is constantly updating itself. This methodology is applicable to a general class of nonlinear systems that are affine in the control with uncertain parameters appearing linearly. Fault-tolerance capability is introduced by incorporating a failure model in the controller so that a failure can be identified and compensated for by a change in the parameters of the failure model. First, model reference adaptive control, structured model reference adaptive control, and structured adaptive model inversion