Abstract: Wilbur Zelinsky and Barrett Lee (1998) recently unveiled a model of the sociospatial process of immigrant settlement designed to augment and possibly supplant the well-known theories of assimilation and pluralism.  Although in some ways new, their work continues a tradition in social science that treats the settlement geography of immigrants as a measure of their more general fit into American society. Using their paper as a springboard, we question the prevailing assumption that immigrant’s settlement patterns represent a barometer of their adaptation, or lack thereof, to a host society. This critique of the concepts of assimilation, pluralism, and Zelinsky and Lee’s alternative, ‘heterolocal’, model of immigrant settlement pivots around the issues of spatial scale and race.  We argue that the contestations over immigration and how well immigrants fit in society are increasingly constructed at the regional scale. We also assert that race questions infuse almost all aspects of these debates. The transformation of America’s largest city-regions into places of non-white immigrants, and the shifting political balance of power to states like California through immigration-driven reapportionment, are touchstones for anti-immigration initiatives and associated local and national debate. Fear of racial regional changes underpins an increasingly powerful response to immigration. The reactions elicited by these settlement geographies fall under the heading we call the ‘territorial politics of immigration’