The U.S. Special Operations Command (USSOCOM) was concerned that decision aids, especially computer-driven models, used to support resource allocation did not adequately represent the contributions of special operations forces (SOF). On the basis of a survey of existing models, RAND concluded that the SOF contribution is not adequately represented. Moreover, many important SOF tasks cannot be captured by models because they are either dominated by uncertainty or are poorly understood. In addition, the doctrinal missions for SOF are just broad areas of employment that do not offer a useful framework for analyzing special operations. These missions should, instead, be analyzed using a hierarchy of objectives descending from national goals to operational tasks and their associated employment concepts. This framework should be applied in four contexts: national-level tasking, large-force operations, guerrilla warfare, and other use, capturing the collateral employment of SOF. Each context differs with respect to typical objectives, command and control arrangements, and characteristic missions. For example, a national-level tasking is set by the national command authority, is controlled at a very high level, and often involves counterterrorism, although SOF might perform any other mission in this context. The study concludes with broad recommendations for a cautious approach to modeling and a briefly stated rationale for maintaining an elite SOF that generates little sustained combat power. This rationale includes leverage, unique capabilities, audacity, flexibility, low visibility, and guerrilla warfare, in which SOF excel.