Seismic hazard assessments in the central and eastern United States (CEUS) usually rely heavily on the historical seismicity record to define the location of earthquake sources as well as the recurrence rate of these sources. Therefore, the stationarity of earthquake activity in time and space is very important to assessing the likelihood of future earthquakes at various locations. Earthquake recurrence intervals, which are on the order of hundreds to thousands of years, have been developed from paleoseismic data at New Madrid, Charleston, Charlevoix, and the Meers fault. The recurrence rates imply cumulative Quaternary displacements that are much larger than have been observed. One explanation for this discrepancy is temporal clustering of earthquake activity whereby active clusters are separated by long periods of quiescence. Available data suggest that active clusters consist of two or more earthquakes and last for longer than several hundred to a few thousand years. The spatial extent of active dusters in the CEUS is the subject of ongoing studies and preliminary data suggest that the active zone is generally in the region of the large observed earthquakes and on-going smaller magnitude seismicity. However, the historical seismic quiescence of the Meers fault, despite geologic evidence for recent activity, points to the need for geologic studies to supplement seismicity data in identifying seismic sources and evaluating their earthquake potential.