The purpose of this study is to optimize adsorption conditions of powdered alum sludge (PAS) as low-cost adsorbent for the removal of three P-species (ortho-P, poly-P and organic-P) from wastewater using the response surface methodology (RSM). Initially, RSM in the basis of a 3-variable Box-Behnken design was used to determine the effect of pH (from 4 to 7), PAS mass (from 0.1 to 0.5 g) and PAS particle size (from 125 to 420 mu m) on the response levels (removal efficiencies of the three P-species). Three response surface quadratic models in terms of three factors were then obtained from an analysis of the experimental data using a SAS computer package. Thereafter, the effect of each of the parameters on P removal for each of the three species was examined using the three-dimensional response surface. All three parameters (pH, PAS mass and PAS particle size) had a significant effect on the removal of each of the P species. Finally, optimal conditions for P species removal were determined at which the P-removals of 99.8% (for ortho-P), 94.9% (for poly-P) and 94.8% (for organic-P) were achieved, respectively. The results derived from the verification experiments agreed with that predicted by the models, confirming the suitability of the established models and the success of RSM in optimizing the PAS adsorption conditions.