This paper presents a novel signal processing method which utilizes the wavelets theory. In the proposed procedure not known from the literature mother wavelet is applied but, through the intensive exploitation of the knowledge of the ideal shape in the considered problem, a new function is defined. In the work, the peak-shaped mother wavelet dedicated for voltammetry was defined and applied to resolve different interpretation problems, such as denoising, baseline correction and overlapping peak separation. The features of the signals were extracted and analyzed in the wavelet coefficients space by use of continuous wavelet transform. Satisfactory results were obtained for simulated voltammograms disrupted by noises or other superimposed undesired signals. Additionally, the process of whole quantitative determination was presented for the laboratory sample containing thallium and lead. The proposed approach is multipurpose and may be applied to process different signals in the case when the shape of an ideal curve is known, can be measured, calculated or obtained by way of simulation. (c) 2007 Elsevier B.V. All rights reserved.