Advocates for older central cities increasingly refer to them as legacy cities in an attempt to shift policy discussions from their social liabilities to their economic assets. In examining this concept, we place U.S. central cities within the context of their regions and use cluster-discriminant analysis to distinguish between groups of regional economies. The cluster analysis subsets metropolitan areas according to their shared position on several theory-driven dimensions. The discriminate analysis identifies the variables that are most closely associated with the statistical grouping of each subset. The results are used to answer a series of questions: 1) Is population size a good proxy for agglomeration? 2) Do small legacy regions exist? 3) Does the typology of legacy regions provide insights about the economic performance of the clusters? Our findings suggest that the economic development paths of regional economies have to change before the future of central cities that experienced severe population losses can shift to a positive trajectory.