Large processes such as globalization, migration and the expansion and further integration of the European Union have led to an increased awareness of issues of collective identity in contemporary Europe. In the literature several competing claims can be found. Some observers discern a trend towards the emergence of a quasi-national European identity and an eclipse of national identities; others see a revival of nationalist sentiments such as ethnocentrism and xenophobia, whereas still others emphasize the growing importance of local and regional affiliations. Using data from the European Values Study and the Eurobarometer, this article attempts to test these competing claims. Our descriptive results show that the current claims are too uniform (there are considerable cross-national differences within Europe) and rectilinear (there are no clear over-time patterns discernible). To gain an understanding of why the somewhat confusing results are as they are, we look, with the assistance of ethnic competition theory, at the underlying causal mechanism at work at both the country and the individual level. Within-country differences in people's geographic identity and their nationalist attitudes appear to be much larger than between country differences, i.e. differences between countries are much more determined by the composition (in terms of gender, age, income, education and religion) of their population, than in terms of context (ethnic, language and religious fractionalization, prosperity and economic growth and immigration). Men, the young, those with higher income, higher education, and the non-religious identify with a broader geographical unit. The older people are and the less educated they are, the prouder they are of their country. The higher the language and religious fractionalization of a country, the lower the level of patriotism. The higher the prosperity of a country, the lower the chauvinism (ethnocentrism and xenophobia); young countries, those with higher incomes and higher education are less chauvinistic.