Sustainable energy development worldwide requires planners and analysts to anticipate possible future outcomes in a world filled with uncertainty. Computer models represent a critical tool that can be used to examine how energy systems might evolve over time under different assumptions. Energy system models can address questions such as “How might the current 23% share of coal in the U.S. energy system change under future climate or energy policies?” The models consider resource availability and pricing, technology innovation, and changes in demand over time. Despite the critical importance of such questions to decision makers, there is little formal guidance regarding how best to develop and apply energy system models in a way that yields useful insight. Dr. Joe DeCarolis, who employs energy system models in his research program, spent a month this past summer as a fellow at University College London (UCL) in their Whole Systems Energy Modelling (wholeSEM) consortium. DeCarolis collaborated with a team of UCL researchers to outline a set of guiding principles for model-based analysis, including approaches to deal with future uncertainty and ways to communicate model-based results to decision makers.