The performance of a PEM fuel cell depends on the proper design of all the components such as the flow field design in the graphite bipolar plates, the diffusion backing and the MEA. The performance of the MEA depends on both the catalyst material and the support structure. The voltage loss that occurs in a fuel cell can be analyzed as a combination of the following major loss components: (i) losses due to cathode overpotential or sluggish cathode kinetics, (ii) ohmic losses and (iii) mass transport related losses. While the voltage loss due to the cathode kinetics can be substantially improved only through the development of newer catalyst, there is still a possibility of improving the mass transport related losses significantly. Further, if the mass transport losses can be reduced at lower catalyst loadings then that can lead to fuel cells that will meet the current automotive targets. In this talk we will present a model-based optimization approach that targets the problem of reducing the mass transport related losses. A detailed mathematical model that includes the effect of liquid water in all the layers of the fuel cell is used in the optimization. The gains that can be achieved through the development of highly optimized MEAs will be outlined.