Polymer Electrolyte Membrane Fuel Cell (PEM-FC) systems have great promise for providing efficient environmentally acceptable portable power sources for transportation and related applications. However, despite quite significant investments in research and development, current systems lead to inadequate performance in the most important applications and the rate of progress is too slow. An alternative paradigm for attacking this problem was proposed in which theory and simulation were used to predict from first principles the structural, transport, and electronic properties of full PEM-FC systems, allowing the materials and configurations to be optimized computationally prior to experimental synthesis and characterization. The ReaxFF reactive force field was developed to describe complex reactions (including catalysis) nearly as accurately as quantum mechanics but at costs comparable to force field based molecular dynamics. The theory was already validated and simulations reproduced some of the known properties of PEM-FC systems by simulating well-studied systems (e.g., Nafion membrane and Pt catalysts for anode and cathode). Thus, the theory and simulation could be used to predict the properties of new systems in which catalysts are designed to provide lower barriers (overpotentials) with reduced effects of poisoning and in which the membranes are modified to optimize proton transport at higher temperatures while minimizing transport of fuel, oxidant and products. This approach enabled a new strategy for fuel cell development in which the possibility of combinatorial computational design is considered to test many possible alloys and membrane compositions computationally. This is an abstract of a paper presented at the AIChE Annual Meeting and Fall Showcase (Cincinnati, OH 1/04/2005).