Evacuation models, including engineering hand calculations and computational tools, are used to calculate the time it takes to evacuate a building, which can then be used in an engineering safety analysis. However, there is a lack of available data and theory on occupant behavior for use by evacuation models to estimate evacuation time results and their uncertainty. In lieu of data and theory, evacuation models (and users) make assumptions and simplifications about occupant behavior, which can inappropriately characterize the time it actually takes to evacuate a building. The purpose of this paper is to reevaluate current egress modeling techniques and advocate for the inclusion of a robust, comprehensive, and validated conceptual model of occupant behavior during building fires. This paper begins by describing the current state of evacuation modeling of human behavior in fires and identifying gaps in current behavioral prediction techniques. The second part of the paper outlines a model of occupant decision-making during emergencies, referred to as the protective action decision model (PADM); a theory that can serve as the basis for the development of a conceptual model of occupant decision-making and behavior during the pre-evacuation period of building fires. The PADM provides a framework that describes the decision-making steps that influence protective actions taken in response to natural and technological disasters—including perceiving information, paying attention to the information, comprehending the information, establishing the nature of the threat, personalizing the risk, searching for potential protective actions and choosing one of these, and then performing that action. The paper ends with a discussion of how to adapt and expand the PADM in order to develop a predictive conceptual model of the pre-evacuation period for use by computer evacuation models.