Polycentricity is often conceived to be the product of “a long process of very extended decentralization from big central cities to adjacent smaller ones, old and new” (Hall and Pain, 2006: 3). Accordingly, polycentric spatial development (as a process) is usually identified with a development towards a more balanced distribution of functions and activities across regional space, and with the moderation of intra-regional urban hierarchies. However, the term polycentricity is also linked to the idea that formerly independent, but close-by and well–linked cities start to ‘fuse’ into larger metropolitan areas as their spheres of influence start to interfere. Classic products of this kind of polycentric spatial development can be found in Europe (e.g. the Randstad), but also in the U.S. such areas have been identified (e.g. SF bay Area, Dallas-Fort Worth, Baltimore-Washington, Raleigh-Durham etc). While ‘fusion-mode’ and ‘decentralization-mode’ metropolitan areas are subject to grossly similar overarching trends such as globalization and the postindustrialization and informationalization of the economy, these trends seem to produce contrasting patterns of spatial organization. The dominant trend in ‘fusion-mode’ polycentric metropolitan areas appears to be towards the creation of new leading cities and a strengthening rather than a moderation of intra-regional hierarchies. In this paper, we explore this trend for ‘fusion-mode’ polycentric metropolitan areas in the U.S., thereby employing occupational data, and seek to explain the emerging new intra-regional hierarchies found.