A learning curve model has been developed to analyze the mass production cost structure of proton exchange membrane fuel cells for automobiles. The fuel cell stack cost is aggregated by the cost of membranes, platinum, electrodes, bipolar plates, peripherals and assembly process. The mass production effects on these components are estimated. Nine scenarios with different progress ratios and future power densities are calculated by the learning curve for cumulative production of 50 000 and 5 million vehicles. The results showed that the fuel cell stack cost could be reduced to the same level as that of an internal combustion engine today, and that the key factors are power density improvement and mass production process of bipolar plates and electrodes for reducing total cost of fuel cell stack. (C) 2003 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.