The quality of electricity has been gaining more emphasis among utilities, service sectors and consumers. They have to maintain the quality by strategic measures in coping with all sort of disturbances generated intrinsically in modern power electronic equipments, large commercial buildings and open power environment. A means of improving electric power quality starts by a systematic identification of the power system disturbances, which is posed to be a big challenge. The conventional approach based on Fourier transform principles has its main drawback of losing the time-domain feature after transformation. Whist the technique of using wavelet transform appears to be more promising with its strength on handling signals on short time intervals for high frequency components and long time intervals for low frequency components. In this paper, an integrated approach, using both Fourier and wavelet transforms, is proposed and it is used to integrate the advantages of both transforms. The wavelet transform is used to extract the required time-domain information from the high frequency components while the Fourier transform is used to provide the accurate measurement from the low frequency components. An automatic power quality recognition system based on the integrated approach is developed. Neural network classifier and rule-based classifier are selected to implement the proposed approach of which its validation is performed via simulated data set.