Distributed interactive simulation (DIS) is a revolutionary development that is changing permanently many features of Defense Department work. The field of warfare modeling and simulation (M&S) is quite large, however, and this paper reviews its evolution and how DIS fits within it. A few things are clear. First, DIS is exceedingly valuable for training. Second, DIS could be a powerful means for improving the quality of planning and analysis if used wisely for occasional well designed experiments. These could 1) provide insights about real-world processes involving human performance and behavior (including decision making), 2) help inform and calibrate models, and 3) help test plans in a quasioperational environment. Third, distributed war gaming, which depends on DIS technology, is already lowering boundaries among developers, planners, and warfighters. This can shorten development processes and improve the results through virtual prototyping; it can also improve operational readiness. To achieve these potential benefits, however, will require a holistic approach, conceptual breakthroughs, and profoundly difficult model development efforts. The challenges include developing integrated hierarchies of models, developing adaptive decision models and other models of human behavior, developing and using new forms of uncertainty-sensitive analysis, and learning how to use DIS experiments effectively.>