Seismic hazard zoning maps are a relatively simple and transparent consequence of the patterns of historical seismicity. Over time, we have refined our understanding of where and why earthquakes occur in continental interiors, and we have improved our characterization of the resulting ground motions and their probabilities. We have begun to understand the important role of uncertainty in seismic hazard analysis. However, there are still significant shortcomings in our treatment of uncertainty that are particularly pronounced for midplate regions. The same lack of knowledge that causes our uncertainty of the hazard also prevents us from accurately quantifying that uncertainty. Are the resulting seismic hazard maps reasonable for use in building codes, in light of this uncertainty?

In order to address this, a simple probability-based areal test of seismic hazard maps can be conducted that employs correlations between modified Mercalli intensity and ground-motion amplitude, in combination with the historic record of seismicity over the past 200 yr. This test shows that hazard maps are reflecting the potential for repeats of the largest historical events in areas that currently experience moderate seismicity. In any 50 yr time period, these areas will experience or exceed the expected area for damaging ground motions if there has been at least one event of M > 6.7 anywhere in eastern North America (or two events of M 6.5). If there have not been any large events, the damage area will be much less than predicted.