A general strategy for the exploration of crystalline rock masses in the eastern United States for the identification of potential sites for high-level radioactive waste repositories has been generated by consideration of the Department of Energy (DOE) Siting Guidelines, available information on these crystalline rocks, and the capabilities and limitations of various exploration methods. The DOE has recently screened over 200 crystalline rock masses in 17 states by means of literature surveys and has recommended 12 rock masses for more intensive investigation including field investigations. The suggested strategy applies to the next stage of screening where the objective is to identify those potential sites that merit detailed site characterization including an exploratory shaft and underground study. The DOE Siting Guidelines are reviewed to determine the types of information that are both important and suitable for collection prior to site characterization; that is, the area phase of exploration, in the terminology used by DOE in the screening process. The single most critical need for information concerns the issue of whether or not the hydrology of a potential site can be adequately characterized. There is almost no information on hydrology at repository depths in the areas being screened by DOE. Early in the exploration of an area, a preliminary delineation of the regional and local ground water flow systems should be made. The locations of fracture zones, faults, shear zones, large dikes, large inclusions of country rock in intrusive rocks, and major changes of rock type need to be determined to begin to characterize the hydro!ogic system and to avoid such features, if possible, in siting the repository. Major fracture zones are significant conduits for water flow in some areas as shown by a summary of deep holes and excavations in the eastern United States and Canada. Details of these holes and excavations, which unfortunately are very limited in number, are presented in an Appendix. The survey of deep holes shows that subhorizontal fracture zones will probably be present in the subsurface in most areas, but these zones appear to be widely spaced below depths of 200 m. Exploration methods that may be used at the area phase include reconnaissance hydrology, remote sensing, geologic mapping, seismic monitoring, mineral resource studies, potential field geophysics, electrical and electromagnetic techniques, active seismology, drilling, borehole logging, core studies, sampling for hydrochemistry, hydrologic testing, and in situ stress measurements. Three phases of exploration are suggested: 1) reconnaissance, which makes use of existing information and includes reconnaissance hydrology, remote sensing, and potential-field geophysics; 2) surface study, which involves detailed ground studies of low cost, such as surface water and spring sampling, geologic mapping, potential-field geophysics, electromagnetic methods, and mineral resource potential studies; and 3) drilling, which directly investigates hydrologic c...