A new algorithm was developed for arcing fault detection based on high-frequency current transients analyzed with wavelet transforms to avoid automatic reclosing on permanent faults. The characteristics of arc currents during transient faults were investigated. The current curves of transient and permanent faults are quite similar since current variation from the fault arc is much less than the voltage variation. However, the fault current details are quite different because of the arc extinguishing and reigniting. Dyadic wavelet transforms were used to identify the current variation since wavelet transform has time-frequency localization ability. Many electric magnetic transient program (EMTP) simulations have verified the feasibility of the algorithm. A new algorithm was developed for arcing fault detection based on high-frequency current transients analyzed with wavelet transforms to avoid automatic reclosing on permanent faults. The characteristics of arc currents during transient faults were investigated. since current variation from the fault arc is much less than the voltage variation. However, the fault current details are quite different because of the arc extinguishing and reigniting. Dyadic wavelet transforms were used to identify the current variation since wavelet transform has time-frequency localization ability. Many electric magnetic transient program (EMTP) simulations have verified the feasibility of the algorithm.