Forest ecosystems are distinguished by their complexity and diversity, their progressive nature of development, and the constant risk of random catastrophes. These systems develop relatively slowly, and research conducted on them requires years or decades to yield meaningful findings with broad extension. Prolonged development, coupled with uncertainty about disruption, work against long-term commitment to fund and maintain study sites. Alternative research strategies (chronosequence and retrospective studies) produce rapid results, but are marred by limited application and uncertainty about past events. With notable exceptions, most long-term studies have centered on relatively undisturbed natural ecosystems. While such studies are useful, they lack the flexibility needed to gauge the long-term impacts of manipulation. New programs, such as the National Science Foundation's network of Long-Term Ecological Research sites, and the USDA Forest Service's Long-Term Soil Productivity studies, encompass and support a wider range of objectives and variables than was possible in earlier studies.