An indigenous purple nonsulfur bacterium Rhodopseudomonas palustris WP3-5 was used to produce hydrogen via photo-fermentation. Response surface methodology was used to optimize the concentration of three critical medium components (butyric acid (HBu), glutamic acid, and FeCl3) having major impacts on the cell growth and H-2 production of R. palustris WP3-5. Four performance indexes (PIs), namely, H-2 yield (Y-H2), maximum H-2 production rate (R-max), overall H-2 production rate (R-overall), and lag time () were used to assess the effectiveness of the phototrophic H-2 production. A new overlay 3-D contour surface plot was developed to determine optimal conditions for R-max, R-overall, and in the system, while maintaining a high H-2 Yield (Y-H2 = 5.5 mol H-2/mol HBu) simultaneously. The overlay optimal regions of the three double responses (Y-H2 - R-max, Y-H2 - R-overall, and Y-H2 - lambda) were identified. Using R-max as the PI, the response surface analysis attained an optimal concentration of 1832, 607, and 54 mg/l for butyric acid, glutamic acid, and FeCl3, respectively, giving an expected R-max and Y-H2 value of 24.9 ml/h/l and 5.74 mol H-2/mol HBu, respectively. This optimal performance is superior to most of the reported values in the literature, indicating that the statistical experimental design was an effective tool leading to marked improvement in phototrophic H-2 production with the R. palustris WP3-5 strain. (C) 2006 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.