A proposed intelligent control system (ICS) contains a sensor fault detection, isolation, and accommodation scheme. If the differences between the estimated and sensed system outputs exceed some threshold values, fault detection logic will initiate a parameter estimation algorithm which determines the type and magnitude of the failure. This is achieved through the use of fault parameters, variables defined to convert the model from that of the unimpaired system to that of an impaired one. Once the fault is isolated, the control system will accommodate it, if possible. This Note discusses the technique used for detecting, isolating, and identifying the fault. The test bed for this research is the T700 turboshaft engine. In the simplified model used here there is one input, fuel flow Wf, and four measured variables, gas generator speed N,, inter-turbine gas temperature T(sub 45), inter-turbine gas pressure P(sub 45), and power turbine torque output Q(sub PT). Should any of the sensors fail, the engine would appear to be malfunctioning and, if the faulty measurement were fed back through the control system, the engine would operate off the design point.