Quantitative differentiation of the various oxidation states of chromium by coprecipitation and preconcentration of trivalent chromium in aqueous solutions using iron(III) hydroxide as the coprecipitating agent has been studied. Recoveries of trivalent chromium are less than 100% because of formation of complexes of trivalent chromium with organic ligands in model solutions. To overcome this problem, the standard addition method has been used. The total chromium concentration in the model solution can be directly determined by graphite furnace atomic absorption spectrometry since the technique has the same sensitivity for both oxidation states of chromium. Because of low concentrations of chromium in the aquatic environment, direct determination of the total chromium concentration in natural waters by graphite furnace atomic absorption spectrometry would require higher sensitivity than the technique is now capable of providing. To improve the sensitivity and selectivity of graphite furnace atomic absorption spectrometry, studies have been made on the fundamental aspects of atomization of chromium in electrothermal atomizers, the effect of heating rates on the sensitivity, the effect of atomization surface on the atomization, and the mechanism of atomization. These studies have provided an insight into atomization mechanisms for chromium and an understanding of the factors responsible for sensitivity and selectivity in analysis. Furthermore, interconvertibleity of the two oxidation states of chromium in the presence of humic substances in aqueous solutions has been studied. The results indicate that when humic acid or fulvic acid is present, the amount of Cr(VI) reduced to Cr(III) is both time- and pH-dependent.