Abstract: Aiming at the actuality that Ion Mobility Spectrometry (IMS) signal is disturbed by the noise, the wavelet threshold denoising method is developed for decreasing the original IMS signal noise. Taking 1ng TNT signals in the image as trial objects, we established a wavelet transform based denoising model by optimizing some key parameters, including threshold calculation principle, threshold processing methods, and wavelet selection. The result showed that the noise of original IMS signal was suppressed effectively according to the universal threshold selection rule after we decomposed the original signal in three layers with sym5 wavelet and processed and reconstructed the wavelet coefficients with the soft threshold method.