The selection of forested natural areas for research and educational p
urposes is discussed. Five factors are important: sufficient size; rep
resentation of typical communities and sites; documented disturbance h
istories; acceptable current condition in terms of age, tree size, and
successional stage; and administrative feasibility.