In order to monitor operating performance of contaminated outdoor insulators, a wavelet-based flashover prediction method is proposed. In most cases, the low-frequency components, namely, fundamental, $3^{rd}$, and $5^{th}$ harmonic components have been mainly used for the sake of the spectral analysis of the leakage current. However, in this paper, the detail coefficients of wavelet transform representing high-frequency components are used as important information to predict a flashover in the contaminated insulator. Experimental results verify that the proposed method has a superior capability for flashover prediction.