Despite the dedicated search for novel catalysts for fuel cell applications, the intrinsic oxygen reduction reaction (ORR) activity of materials has not improved significantly over the past decade. Here, we review the role of theory in understanding the ORR mechanism and highlight the descriptor-based approaches that have been used to identify catalysts with increased activity. Specifically, by showing that the performance of the commonly studied materials (e.g., metals, alloys, carbons, etc.) is limited by unfavorable scaling relationships (for binding energies of reaction intermediates), we present a number of alternative strategies that may lead to the design and discovery of more promising materials for ORR.