Engineering strategies were applied to promote the phototrophic H-2 production of an indigenous purple nonsulfur bacterium Rhodopseudomonas palustris WP3-5 using major components (i.e., acetate, butyrate, and lactate) of dark fermentation effluents as carbon sources. First, performance of cell growth and photo-H-2 production on each carbon source was examined individually. It appeared that acetate was the most effective carbon source for photo-H-2 production, giving an overall H-2 production rate and H-2 yield of 12.68 ml/h/1 and 67.1%, respectively. Next, the effect of substrate concentration of each carbon source on photo-hydrogen production was investigated. Kinetic models were developed to describe the correlation between maximum specific growth rate/specific H-2 production rate and the substrate concentration. The results show that using acetate and lactate as the carbon source, the kinetics for the cell growth and photo-hydrogen production can be described by Monod-type and Michaelis-Menten models, respectively, whereas substrate inhibition occurred when using butyrate as the carbon source. The continuous cultures were also conducted at a hydraulic retention time of 96 h using synthetic dark fermentation soluble metabolites (with a 5 and 10 fold dilution) as the influent. The phototrophic H-2 production efficiency was stably maintained for over 30 days with an overall H-2 yield 10.30 and 11.97 mol H-2/mol sucrose, when using 5-fold and 10-fold diluted dark fermentation effluent, respectively, as the substrate for dark fermentation. This demonstrates the feasibility of using the sequential dark and photo fermentation for high-yield biohydrogen production. Copyright (C) 2011, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.