Sucrose and pineapple waste (PAW) were used as the carbon substrate to produce H, using municipal sewage sludge as the H-2 producer. The effects of pH, substrate concentration and type of buffer on H-2 production were studied. Regardless of the type of carbon substrate, phosphate-buffered medium exhibited much higher H-2 production efficiency and lower CO2/H-2 ratio than those obtained in medium containing bicarbonate buffer. For H-2 production from sucrose-based medium, the best pH and carbon substrate (sucrose) concentration was 7.5 and 20 g COD/l, respectively, giving a maximum H-2 production rate (v(H2,max)) and H-2 yield (Y-H2) of 745 ml/h/l and 2.46 mol H-2/mol sucrose (or 5.27 mol H-2/ g COD), respectively. When using PAW as the substrate, the H-2 production performance was similar with or without additional nitrogen source, indicating that the PAW substrate contained sufficient nitrogen source for cell growth and H-2 production. The v(H2,max) and YH2 obtained from phosphate-buffered PAW medium (Medium B) was 383 ml/h/l and 5.92 mol H-2/g COD, respectively, which is 95 and 64% higher than the carbonate-buffered medium (Medium D). Although attaining lower H, production rate than sucrose medium, the PAW-based medium gave comparable H, yield, suggesting the feasibility of using the renewable resource (PAW waste) for fermentative H, production. Increasing the culture volume (using PAW medium) from 100 ml to 2.5 L (a 25-fold increase) caused a 20-40% decrease in the H, production rate and H-2 yield, probably due to poor mixing efficiency in the scale-up experiments. (c) 2006 Elsevier Ltd. All rights reserved.