Smoke from wildfires has adverse biological and social consequences, and various lines of evidence suggest that smoke from wildfires in the future may be more intense and widespread, demanding that methods be developed to address its effects on people, ecosystems, and the atmosphere. In this paper, we present the essential ingredients of a modeling system for projecting smoke consequences in a rapidly warming climate that is expected to change wildfire regimes significantly. We describe each component of the system, offer suggestions for the elements of a modeling agenda, and provide some general guidelines for making choices among potential components. We address a prospective audience of researchers whom we expect to be fluent already in building some or many of these components, so we neither prescribe nor advocate particular models or software. Instead, our intent is to highlight fruitful ways of thinking about the task as a whole and its components, while providing substantial, if not exhaustive, documentation from the primary literature as reference. This paper provides a guide to the complexities of smoke modeling under climate change, and a research agenda for developing a modeling system that is equal to the task while being feasible with current resources.