Keywords:As(V) As(III) Sorption Chitosan Chitosan-immobilized sodium silicate a b s t r a c t A natural biosorbent containing amine functional groups, chitosan, and a novel sorbent, chitosanimmobilized sodium silicate, were prepared and utilized for the selective sorption of As(V) from waters prior to its determination by atomic spectrometric techniques, namely, hydride generation atomic absorption spectrometry (HGAAS) and inductively coupled plasma mass spectrometry (ICP-MS). Chitosan was synthesized from chitin and sodium silicate was used as the immobilization matrix due to its straightforward synthesis. Through sequential sorption studies, it was shown that chitosan-immobilized sodium silicate has exhibited a better chemical stability than the chitosan itself which demonstrates the advantage of immobilization method. Both chitosan and chitosan-immobilized sodium silicate were shown to selectively adsorb As(V), arsenate, from waters at pH 3.0 at which neither chitin nor sodium silicate displayed any sorption towards As(V). The sorption of arsenate by chitosan is supposed to have electrostatic nature since pH of 3.0 is both the point at which the amino groups in chitosan are protonated and also the predominant form of As(V) is H 2 AsO 4 − . A pre-oxidation step is required if both As(III) and As(V) are to be determined. Desorption from the sorbents was realized with 1.0% (w/v) l-cysteine prepared in a pH 3.0 solution adjusted with HCl. Among the possible interfering species tested, only Te(IV) and Sb(III) were shown to cause a decrease in the sorption capacity especially at high interferant concentrations. High concentrations of Sb(III) also resulted in gas phase interference during hydride generation.The validity of the method was checked both via spike recovery experiments and also through the analysis of a standard reference material. Spike recovery tests were carried out with four different types of water; namely, ultra-pure, bottled drinking, tap, and sea water; and percent recovery values were found to be 114 (±4), 112 (±2), 43 (±4), and 0 (±1), respectively. It was concluded that the proposed methodology can be applied efficiently to low-to-medium ionic strength solutions, such as most drinking waters. The accuracy of the method was additionally investigated through the analysis of a standard reference material and a good correlation was found between the determined (26.6 ± 2.4 g L −1 ) and the certified (26.67 g L −1 ) value.