A diagnostic process capable of providing an early warning of a fault in a gas turbine is of tremendous value to the user and can result in substantial financial savings. The approach in the Genetic Algorithm based technique adopted is to treat the problem of engine diagnostics as an optimisation exercise using sensor-based and mathematical behavioural model based information. The engine performance model would simulate a range of possible combinations of potential faults (i.e the effects of model-based information) and a comparison would be made with values of the actual (sensor-based) parameters obtained from an engine. The difference between the actual and simulated values of would be converted into a suitable objective-function and the aim of the optimisation technique such as the genetic algorithm would be to minimise the objective function. The technique has given promising results for simple cycle engines.