Uncertainty is a pervasive feature in natural resource management. Adaptive management, an approach that focuses on identifying critical uncertainties to be reduced via diagnostic management experiments, is one favored approach for tackling this reality. While adaptive management is identified as a key method in the environmental management toolbox, there remains a lack of clarity over when its use is appropriate or feasible. Its implementation is often viewed as suitable only in a limited set of circumstances. Here we restructure some of the ideas supporting this view, and show why much of the pessimism around AM may be unwarranted. We present a new framework for deciding when AM is appropriate, feasible, and subsequently successful. We thus present a new paradigm for adaptive management that shows that there are no categorical limitations to its appropriate use, the boundaries of application being defined by problem conception and the resources available to managers. In doing so we also separate adaptive management as a management tool, from the burden of failures that result from the complex policy, social, and institutional environment within which management occurs.