Surface subsidence is an inevitable consequence of most underground coal mining. The United States Bureau of Mines has estimated that over 3,000 square miles of land in this country have been affected by subsidence, and it is currently estimated that an additional 2,400 to 3,800 square miles will be affected over the next twenty years. Surface subsidence is a complex phenomenon which is influenced by many variables related to both mining and site conditions. Although it cannot be prevented, foreigh experience has shown that it, nevertheless, can be controlled. In order to achieve this goal, however, methods of subsidence prediction and control must be developed for the United States mining conditions and justified with empirical data. The objective of this research is to develop characteristic relationships of longwall subsidence and its related parameters from case studies gathered from the Appalachian coalfields. Furthermore, based on this information, an empirical, predictive capability will be developed which can be used to provide accurate and reliable predictions of subsidence in this region. / Master of Science