The Chaos paradigm focuses on patterns and the sometimes complex dependency of outcomes on initial conditions. This paradigm has been increasingly used across a spectrum of physical science disciplines as scientists have become aware that physical reality is not best characterized as a group of isolated events in which steady states are altered in a predictable, systematic manner by interventions. A framework for prevention research designed upon the chaos paradigm has a number of characteristics. It views individuals as "actors" rather than "subjects," who are situated within a number of complex, interwoven, dynamic systems. Also, no optimal "ideal" state of equilibrium is recognized as the norm, but rather a variety of possible equilibrations, each with its own continuing history and own set of advantages and disadvantages. To characterize this complexity, a generic relational framework is required to identify "actors" and systems that may be relevant to the outcomes of interest and their relationship to each other regarding those outcomes. This framework must have the flexibility and relativity the ethnographic perspective brings to social units and cultures. Finally, in order to draw general conclusions, the framework must be based on indepth study of a large number of events in which documentation is obtained regarding the initial relationships of actors and relevant systems, how implementation of an intervention affects and is affected by relevant systems, and the effect on a variety of individual and environmental outcomes. A framework and research strategy that incorporates these elements is being developed in a research study that is examining implementation and outcomes across 70 demonstration programs funded by the Center for Substance Abuse Prevention. 21 references