East of the Rocky Mountain front, recurrence intervals of most large earthquakes are long compared to the historical earthquake record. Information about longer-term, prehistoric earthquake histories can reduce the resulting uncertainty in hazard assessments. Accordingly, paleoseismological results and selected aspects of bedrock geology, such as the boundary between the craton and adjacent Phanerozoic terranes, were incorporated into the 1996 USGS national seismic-hazard maps to extend the total record of seismicity back into prehistory. We identify post-1996 findings that should be incorporated during updating of the central U.S. part of the 1996 maps for publication in 2002. A companion paper (Cramer et al., 2002) determines the impact of each new finding on the calculated hazard. We propose three adjustments to the largest moment magnitude, M(max), that is assumed to occur within a specified area. (1) Throughout the slightly active central craton of the central and eastern United States, paleoseismological evidence from the Illinois basin and global seismological evidence require that M(max) be increased above the value of M 6.5 assumed for the 1996 maps, to M 7.0. (2) We recommend increasing M(max) from M 6.5 to M 7.5 within the Rough Creek graben of western Kentucky and southeastern Illinois, the Rome trough in eastern Kentucky, and the Grayville graben of Illinois and Indiana. The increases would achieve geological and logical consistency with M(max) values in other parts of the central and eastern U.S. (3) Two prehistoric earthquakes of estimated M 7.1 and M 7.5 near the Illinois-Indiana border occurred within the craton. Absent evidence to justify raising the cratonic M(max) to 7.5, we specify a zone surrounding the two prehistoric earthquakes where M(max) could be locally raised to M 7.5.