Phototrophic hydrogen production by indigenous purple non-sulfur bacteria, Rhodopseudomonas palustris PBUM001 from palm oil mill effluent (POME) was optimized using response surface methodology (RSM). The process parameters studied include inoculum sizes (% v/v), POME concentration (% v/v), light intensity (klux), agitation (rpm) and pH. The experimental data on cumulative hydrogen production and COD reduction were fitted into a quadratic polynomial model using response surface regression analysis. The path to optimal process conditions was determined by analyzing response surface three-dimensional surface plot and contour plot. Statistical analysis on experimental data collected following Box-Behnken design showed that 100% (v/v) POME concentration, 10% (v/v) inoculum size, light intensity at 4.0 klux, agitation rate at 250 rpm and pH of 6 were the best conditions. The maximum predicted cumulative hydrogen production and COD reduction obtained under these conditions was 1.05 ml H(2)/ml POME and 31.71% respectively. Subsequent verification experiments at optimal process values gave the maximum yield of cumulative hydrogen at 0.66 +/- 0.07 ml H(2)/ml POME and COD reduction at 30.54 +/- 9.85%. (C) 2009 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.