What determines support for European integration? Easton (1975) argues that political support ranges from the specific to the diffuse. Citizens usually assess national governments in terms of short-term performance and judge governmental institutions in terms of long-term experience. European integration is neither a government nor a set of institutions; it is a process The European Union (EU) also does not govern in the traditional sense of the term. Can Easton’s framework be applied here? If so, how are these evaluations framed? The EU will be defined as an asymmetric federation: duration of membership, the degree of shared trade, and financial costs and benefits vary across the member states. These differing contexts could condition how individuals perceive integration. Consequently, this paper will address two questions. First, to what extent do specific and diffuse evaluations drive support for European integration? Second, to what extent are these evaluations framed by the fragmented nature of Europe integration? Data for the analysis will be drawn from responses to identically worded questions in Eurobarometers 51.0, 52.0, 53.0, and 54.1 (which were conducted in the fall and spring of 1999 and 2000). The dependent variable will be an additive index that measures whether respondents prefer national-level or European-level decision-making in 17 major policy areas. Integration increases as national policies are shared at the European level and decreases as European policies are recaptured by the member states. An additive index of support for the four major European-level institutions will represent diffuse support, and an additive index of the perceived national benefits of EU membership will represent specific support. Years of membership, intra-EU trade, and net EU budget contributions will be the key contextual variables. Additional, relevant individual-level and national-level predictors also will be included. When pooled, the surveys include 60 national-level and approximately 58,000 individual-level cases, which permits a multilevel model to be adequately specified at both levels. Consequently, a multilevel rather than a contextual model will be used. This paper should contribute significantly to the literature on European integration. We do not know how either specific or diffuse support contribute to popular perceptions of European integration. Evaluations of integration may be an extension of national-level concerns, as the literature on second-order elections would seem to indicate. Or integration could be assessed via European institutions, in which case some form of European identification potentially may exist. The framing effects of national context are of equal importance. If these perceptions vary with the amount of shared trade or financial contributions, then they may be fleeting; but if they vary with duration of membership, then they may be more enduring. With the introduction of the Euro and the upcoming accession, the EU is at an historical juncture. One only can estimate how far European integration will proceed is one understands how Europeans view this process. [ABSTRACT FROM AUTHOR]