The Department of Defense (DOD) is becoming critically dependent on combat models (including simulations and war games). Substantial interest in improving model-related technology has not been matched by interest in the substance of the models and the validity of the lessons learned from using them. For both research and applications, combat models should be viewed less as answer machines than as frameworks for summarizing and communicating objective and subjective knowledge (including knowledge of uncertainties), and as mechanisms for exploration. Models should also be designed so that they or their modules can be directly compared with, used in, or used with other models. This is crucial for scientific reproducibility and peer review, for efforts to calibrate aggregated models using higher resolution models, for analysis of empirical data, and for distributed war gaming. The authors recommend that the Secretary of Defense establish an Office of Military Science (OMS) to plan and administer the process of creating the national environment necessary for a vigorous military science. The OMS would encourage, nurture, and to some extent sponsor research, although relying primarily on the military services and other agencies for most research and analysis.