BACKGROUND: The sorption of arsenate, a poison of acute toxicity found in natural waters, onto chitosan, a biosorbent derived from waste seafood shells has been studied. A batch adsorber design model was developed to determine how much chitosan adsorbent is required to reduce the arsenate concentration in solutions to the WHO standard of 10 mu g L-1. RESULTS: A series of batch kinetic experiments has been carried out at different initial pH values. The initial arsenate sorption appears to be completed after 30 min, however, a steady reversible reaction takes place resulting in the desorption of arsenate over 48 h. These phenomena in the batch kinetic data have been correlated simultaneously using the newly developed pseudo-first order reversible model. Two batch reactor design models have been developed and compared. The first model is a conventional approach based on the equilibrium isotherm capacity equation. A second batch adsorption reactor design is based on the principle of contacting time required, t(max), for the chitosan to achieve its maximum adsorption capacity, q(max). The practical outcome from the second batch adsorber model results in a saving in adsorbent mass per batch of approximately 39.4%, 96.2% and 92.3% chitosan adsorbent at pH conditions of 3.5, 4.0 and 5.0, respectively. CONCLUSION: The adsorbent cost and handling costs are reduced in the second batch adsorber model. There is also a significant savings in the batch turnaround time required in the batch adsorber design when the design is based on the maximum adsorption capacity rather than the equilibrium adsorption capacity. (C) 2010 Society of Chemical Industry