This article analyses the changes in industry concentration in the European Union (EU) using two regional samples: the EU-15 for the years 1980-2004 and the Central/Eastern European Countries (CEECs) for the years 1990-2004. We calculate Brülhart and Traeger’s entropy index (2005) for agriculture, manufacturing, and services. To assess the statistical significance of the gaps between sub-periods, we use a block-bootstrap procedure. Our results show a change in industry concentration patterns in the EU-15 and the CEECs, corroborating the findings of economic-geography models, namely, that the decline in transaction costs alters economic concentration patterns. However, those findings also show that an initial concentration phase is followed by a second phase characterised by redispersion at high levels of economic integration. Our results do not confirm that theoretical conclusion. Does this mean that integration has not yet made sufficient progress? To answer this question, we need to estimate transaction costs and their changes as defined in economic-geography models. These estimations are in progress. Moreover, our results reveal that, while concentration patterns are changing, they are not converging between Western and Eastern European countries. This phenomenon is not surprising, given the specific economic characteristics of CEECs compared with the EU-15: the structural change under way in the new Member States due to the transition to a market economy is impacting the distribution of their economic activities, thereby distinguishing CEECs from the older Member States. However, this development is not neutral in regard to the EU cohesion-policy goals for 2013. Higher concentration in agriculture and manufacturing in the CEECs could widen per-capita wage gaps in those countries, a trend that would run counter to EU objectives.