This paper presents all artificial neural network (ANN) based approach to tune the parameters of the static var compensator (SVC) reactive power controller over a wide range of typical load model parameters. The gains of PI (proportional integral) based SVC are optimised for typical values of the load voltage characteristics (n(q)) by conventional techniques. Using the generated data, the method of multilayer feed forward ANN with error back propagation training is employed to tune the parameters of the SVC. An ANN tuned SVC controller has been applied to control the reactive power of a variable slip/speed isolated wind-diesel hybrid power system. It is observed that the maximum deviations of all parameters are more for larger values of n(q). It has been shown that initially synchronous generator supplies the reactive power required by the induction generator and/or load, and the latter reactive power is purely supplied by the SVC. (C) 2007 Elsevier Ltd. All rights reserved.