In this paper, the wavelet transform approach will be applied to monitor the power quality disturbances in the power system of the Tainan Science-Based Industrial Park. The wavelet theory provides a unified framework for signal processing applications. By this way, the characteristics of non-stationary disturbances can be more closely monitored. The transient behavior and discontinuities in the signals can be both investigated. For finding out the major reason of the power quality disturbances, the data collected from every monitoring point are analyzed by using the wavelet transform approach. Test results support the feasibility of this approach for the industry power application.