Abstract
In this paper we argue that the various discussions of the regional location behaviour of the multinational firm by the different fields of analysis which deal with these issues are all rather at a tangent to each other. Only economic geography and regional economics discuss firm-location behaviour at the subnational regional level, whereas international trade theory and traditional international business analysis focus only on firm locations at the level of a country. Where subnational regional locations have recently been discussed in international business analysis, this has been done primarily by incorporating the Porter ‘clusters' literature. However, by adopting a transactions-costs approach, we show that such a ‘clusters' concept is unable to distinguish between whether a multinational enterprise should or should not locate in a particular region. In addition, we use this approach to point to directions of research fruitful for reconciling these various different traditions of location analysis.