Component failures occur in many practical systems and may cause performance deterioration and even lead to system instability and catastrophic accidents. There have been many studies in the literature on control of systems with component failures [1]-[4]. In these papers, different design methods including multiple model, switching and tuning designs, fault detection and diagnosis designs, robust control designs and adaptive designs are used. In many applications, failures are uncertain, that is, during system operation, it is not known when, how many components have failed, which component have failed and the extent of failures are also unknown. Adaptive control approaches are useful to handle uncertainties in both system dynamics and component failures and have gained much attention in recent years [5]-[7]. Compared to other approaches, the direct adaptive control approach has the key advantage that it can provide theoretically provable asymptotic tracking in addition to stability, in the presence of large parameter variation and uncertainties. Important results for direct adaptive control of systems with actuator failures exist in [6], [7]. Delay phenomena are frequently encountered in mechanics, physics, applied mathematics, biology, economics and engineering systems. In the presence of time delay, the design of fault tolerant controller becomes more complex. Therefore, the problem of fault tolerant adaptive control of delay systems has received little attention. For example, A direct state feedback adaptive control scheme is introduced in [8] for linear state delay systems with unknown plant dynamics and unknown constant stuck failures. In this dissertation, model reference adaptive control (MRAC) design is developed for state delay systems and in the presence of unknown actuator failures. All parameters of the system are assumed to be unknown. At the first step, the state delay value is considered to be known. By considering the delay value to be known, the additional feedforward controller can be used to compensate the delayed states of the system with delayed states of the reference model. To drive a suitable control design, the two component controller structure introduced in [9] for adaptive control of state delay systems, is combined with the controller structure defined in [7] for compensating actuator failures. Then the state delay value is considered to be unknown and time-varying. To drive a suitable control design in this case, the controller structure is composed of two terms. The first term is defined for compensating actuator failures and the second integral term is used to achieve robustness against unknown delay values. In the next it is shown that this controller structure has robustness with respect to an external bounded disturbance with unknown bounds in addition to multiple unknown time-varying plant delays. In each case the adaptive controller is designed for both relative degree one and relative degree two cases. To the best knowledge of authors, it is the first output-feedback model reference adaptive controller for compensating actuator failures in state delay systems. Finally A state feedback adaptive controller is designed for a more general type of actuator failure which involves both time varying stuck failures and loss of effectiveness failures. Methodology