This paper surveys the development of sensor failure diagnosis in engine digital control systems.First,the early research works on techniques of sensor failure detection and isolation are reviewed.The main disadvantage of model-based methods is high sensitivity to modeling errors which are inevitable because of the complexity of the aircraft engine.Then,the paper considers the necessity of robust failure diagnosis and its development.In view of the fact that the neural networks have been successively applied to control realm and possess themselves of tremendous potentials,the neural-network-based failure diagnosis is recommended as a promising approach in the engine digital control systems.