Current of arcing fault in the low-voltage distribution lines is decomposed by dyadic wavelet transform in this paper, and each scale high-frequency coefficients of wavelet transform of the current waveform are gotten. Compared with the load normal starting and running state, the high-frequency coefficients have obvious characteristic when the arcing fault occurs. Effectively, the characteristic can be applied to the arcing fault detector. Based on the method, current is collected online and analysed by LabVIEW and MATLAB software platform. The result of experiment shows that the method can accurately identify the arc fault on low-voltage distribution lines.