This study begins with the observation that three similar states - Denmark, Norway and Sweden - have had different EC/EU policies, and that one of the foreign policy literature's most interesting approaches, domestic structures analysis, does not shed light on this variation. The goal of the study is to develop an understanding of the different policies by analyzing the question of EC/EU membership using an approach in which issue area, defined in terms of both substance and impact, is linked to policy process and policy choice. Substantive issue area is studied by analyzing parliamentary debates in the three countries. An issue area typology with four substantive categories - economic, political/ policy, international/security, and other — is used to classify arguments made in the debates. The analysis shows that the question was an economic and political issue in Denmark, Norway and Sweden in the 1990s. It was a security issue in Sweden in the 1960s and 1970s. It is argued that this variation helps explain variation in EC/EU policy. In particular, the findings support the argument that the importance of the substantive definition of the EC/EU question is related to the coalitional possibilities it creates. The size of the coalition has, in turn, an impact on policy process and outcome. When the coalitions are large, the process moves more quickly, and parliaments and political parties tend to be consulted or informed rather than active participants. The question of EC/EU membership is also analyzed using an impact typology in which questions are classified as generating one of three types of conflict: none, managed and unmanaged. The EC/EU membership question is classified on the basis of public opinion data and conflict or agreement within political parties. The analysis shows that there were three cases in which the EC/EU question created managed conflict: Denmark and Sweden in the 1960s and Sweden in the 1970s. In other cases, conflict was unmanaged. In an analysis of the importance of variation with regard to type of conflict, it is argued that the data support the hypothesis that the existence of unmanaged conflict is related to decisions to call referenda to decide the membership question. In the face of unmanaged conflict political elites were encouraged to give authority for decision making over to voters. This was, in turn, linked to the emergence of ad hoc organizations dedicated to influencing public opinion for or against membership. This study suggests that the issue areas approach can offer important contributions to the analysis of foreign policy. A challenge for future research is to analyze how issue areas and domestic structure interact to generate policy process and outcome. Central questions should include the relative importance of the two and analyses of the conditions under which one or the other is likely to dominate. digitalisering@umu