It has been previously noted that mixed communities typically produce more power in microbial fuel cells than pure cultures. If true, this has important implications for the design of microbial fuel cells and for studying the process of electron transfer on anode biofilms. To further evaluate this, Geobacter sulfurreducens was grown with acetate as fuel in a continuous flow 'ministack' system in which the carbon cloth anode and cathode were positioned in close proximity, and the cation-selective membrane surface area was maximized in order to overcome some of the electrochemical limitations that were inherent in fuel cells previously employed for the study of pure cultures. Reducing the size of the anode in order to eliminate cathode limitation resulted in maximum current and power densities per m(2) of anode surface of 4.56 A m(-2) and 1.88 W m(-2) respectively. Electron recovery as current from acetate oxidation was c. 100% when oxygen diffusion into the system was minimized. This performance is comparable to the highest levels previously reported for mixed communities in similar microbial fuel cells and slightly higher than the power output of an anaerobic sludge inoculum in the same ministack system. Minimizing the volume of the anode chamber yielded a volumetric power density of 2.15 kW m(-3), which is the highest power density per volume yet reported for a microbial fuel cell. Geobacter sulfurreducens formed relatively uniform biofilms 3-18 mum thick on the carbon cloth anodes. When graphite sticks served as the anode, the current density (3.10 A m(-2)) was somewhat less than with the carbon cloth anodes, but the biofilms were thicker (c. 50 mum) with a more complex pillar and channel structure. These results suggest that the previously observed disparity in power production in pure and mixed culture microbial fuel cell systems can be attributed more to differences in the fuel cell designs than to any inherent superior capability of mixed cultures to produce more power than pure cultures.