Discrete Wavelet Transform (DWT) has been used in the recent yearsin signal processing applications, i.e. filtering and compression. In thecase of denoising because the energy of the noise is spread in the entire wavelet coefficients and it has low amplitude, it can be rejected by thres holding. In this paper, we propose a model to evaluate the influence of the denoising parameters in the quality of the speech signals, by ablind process. We examine the residual signal to establish an objective and blind criteria for selecting the following parameters: base, levels of de composition, rule, and threshold. This model can be applied in anytype of speech signal, no matter its behavior in time and frequency.