Optimizing the production of microporous activated carbon from waste palm shell was done by applying experimental design methodology. The product, palm shell activated carbon was tested for removal of SO(2) gas from flue gas. The activated carbon production was mathematically described as a function of parameters such as flow rate, activation time and activation temperature of carbonization. These parameters were modeled using response surface methodology. The experiments were carried out as a central composite design consisting of 32 experiments. Quadratic models were developed for surface area, total pore volume. and microporosity in term of micropore fraction. The models were used to obtain the optimum process condition for the production of microporous palm shell activated carbon useful for SO(2) removal. The optimized palm shell activated carbon with surface area of 973 m(2)/g, total pore volume of 0.78 cc/g and micropore fraction of 70.5% showed an excellent agreement with the amount predicted by the statistical analysis. Palm shell activated carbon with higher surface area and microporosity fraction showed good adsorption affinity for SO(2) removal. (C) 2008 Elsevier Ltd. All rights reserved.