This paper describes the use of a moving median filter in chemical applications. A probabilistic approach shows the performance in elimination of outliers. The principal advantage over linear filters is the elimination of outliers without significant deformation of the analytical signal. A decrease in level of the normal random noise is also observed. However, the attenuation is lower that obtained by using the moving average with the same smoothing interval size and it is dependent of derivative of the signal. A characteristic of this method is that the peaks are not broadened, however their tops can be cut off.3102281288