The major goal of the research presented here is to test the usefulness of a different way of conceptualizing broad regions of the United States. The census regions are often used to group MSAs for various types of studies. The alternative regions are defined based on access to ocean, Great Lakes, or river ports. The usefulness of this set of regions is compared to that of the census regions using both a dummy variable approach and an index of disparity approach. This paper presents a statistical test of the hypothesis that access to port facilities could and did positively influence urban growth between 1970 and 1990. The rationale for the hypothesis is that the expansion of trade resulting from the North American Free Trade Association, various steps accomplished under GATT and the WTO, and the United States' leading role as a free trade advocate has increased the advantage of expanding economic activity in coastal regions.