A multicomponent background electrolyte (BGE) was developed and its composition optimized using artificial neural networks (ANN). The optimal BGE composition was found to be 90 mm boric acid, 115 mm Tris, and 0.75 mm EDTA (pH 8.4). A separation voltage of 20 kV, 20degreesC and detection at 210 nm were used. The method was applied to characterize several humic acids originating from various countries of the American continent: soil (Argentina), peat (Brazil), leonardite (Guatemala and Mexico) and coal (United States). Comparison with humic acids of International Humic Substances Society (IHSS) standard samples was also done. Well reproducible electropherograms showing a relatively high number of peaks were obtained. Characterization of the samples by elemental analysis and UV spectrophotometry was also done. In spite of the very different origins, the similarities between humic acids are high and by matrix assisted desorption/ionization-time of flight (MALDI-TOF)-mass spectrometry it was shown that most of the m/z patterns are the same in all humic acids. This means that humic acids of different origin have the same structural units or that they contain the same components.