We examine a direct methanol fuel cell (DMFC) model that captures the essence of electrode kinetics and methanol crossover through the polymer electrolyte membrane. Model parameters and key factors for the DMFC model including methanol crossover are identified. Moreover, we establish a relationship between the methanol feed concentration and the power density at a given current density. To gain insight into the effect of the methanol feed concentration, we also performed sensitivity analysis between the cell voltage and the methanol feed concentration. From this sensitivity curve, we are able to identify the optimal feed concentration, which provides the highest power density output for a set value of the current density. Finally, the optimal response of the cell voltage to the change of the feed concentration is studied via dynamic optimization to the differential algebraic equation system. These dynamic optimization results provide a constant feeding strategy that achieves the highest power density at given operating conditions specified by a set current density. (c) 2005 Elsevier Ltd. All rights reserved.