This paper proposes a theory for regional integration based on structural conditions. By applying power transition theory, we are able to account for and anticipate not only the tendency for pairs of countries to participate in integration, but the likely intensity of such integration as well. We find that integration is most likely when there is an asymmetric distribution of power between countries and when they are jointly satisfied after a power transition has occurred. These results hold even after controlling for level of development and trade flows. We test this theory on all politically relevant dyads between 1950 and 1996 using a novel method to measure the level of regional integration. We illustrate the empirical findings through simulations that track the developments leading to Europe's Economic and Monetary Union (EMU).