The Birge-Massart threshold wavelet methods were used to denoise lightning transient electrical signals, compared with the traditional de-noising methods such as Smoothing filter and FIR digital low-pass filter. The Means square error (MSE) and Magnitude error (ME) of simulation signals were calculated to compare the de-noising effect. The double exponential-decay pulse signals and Gaussian white-noise were composed as the simulating signals. The measured signals were also used to de-nosing with traditional and wavelet threshold methods. The results of simulation data and measured ones proved that the Birge-Massart threshold wavelet method was better than the traditional ones.