This paper presents the observations on the bio-removal of arsenic from contaminated water by using Ralstonia eutropha MTCC 2487 and activated carbon in a batch reactor. The effects of agitation time, pH, type of granular activated carbon (GAC) and initial arsenic concentration (As-o) on the % removal of arsenic have been discussed. Under the experimental conditions, optimum removal was obtained at the pH of 6-7 with agitation time of 100 h. The % removal of As(T) increased initially with the increase in As. and after attaining the maximum removal (similar to 86%) at the As,, value of around 15 ppm, it started to decrease. Simultaneous adsorption bioaccumulation (SABA) was observed, when fresh GAC was used as supporting media for bacterial immobilization. In case of SABA, the % removal of As(III) was almost similar (only similar to 1% more) to the additive values of individual removal of As(III) obtained by only adsorption and only bio-adsorption. However, for As(V) the % removal was less (similar to 8%) than the additive value of the individual % removals obtained by only adsorption and bio-adsorption. Percentage removal of Fe, Mn, Cu and Zn were 65.17%, 72.76%, 98.6% and 99.31 %, respectively. Maximum regeneration (similar to 99.4%) of the used bio-adsorbent was achieved by the treatment with 5NH(2)SO(4) followed by 1N NaOH and 30% H2O2 in HNO3. The fitness of the isotherms to predict the specific uptake for bio-adsorption/accumulation process has been found to decrease in the following order: Temkin isotherm > Langmuir isotherm > Freundlich isotherm. For the adsorption process with fresh GAC the corresponding order is Freundlich isotherm > Langmuir isotherm > Temkin isotherm for As(V) and As(T). However, for As(III) it was Langmuir > Temkin > Freundlich. (c) 2007 Elsevier B.V. All rights reserved.