High Impedance Fault (HIF) has long been a challenging problem in network protection. This is because of its difficulty to be detected and discriminated accurately and reliably. In real cases, arcing is usually associated with HIFs. In this paper, arcing currents are produced by setting up a HV experiment. Normal switching transients are also produced by simulation or experiment. They are then analysed based on wavelet transform. It was found by previous researches that db4 is sensitive in detecting high frequency transient. Different features for the fault currents are found and a detection criterion is developed. The simple detection criterion is used to discriminate the high impedance fault from switching transients and normal load current.Index Terms-High impedance fault, wavelet transform, arcing, capacitor switching, magnetizing inrush.I.