In response to increasing care needs, the development of elderly care systems has become a prominent policy concern across the OECD. Since the beginning of the 1990s, the increasing economic constraints led to the rise of policies that focus on the improvement of “coordination” and “integration” of administrations and services – in other words, policies that aim at reducing the fragmentation of health and social care system. As a result, a growing number of studies focus on innovative integrated care schemes and their transferability from one national context to another. Yet they do not systematically investigate the political, organizational, cultural and societal contexts in which they have emerged neither do they link them to the diverse conceptions of integration or coordination carried by the multiple actors concerned. In this paper, we identified three policy-making mechanisms underpinning integration policies at both national and local levels. Theory/Methods: The POLIA-INLOVE comparative research investigates integration/coordination policies in three European countries – France, Sweden and UK. Its analytical framework borrows from sociology, public policy analysis and management studies. Based on a qualitative and diachronic perspective, not presupposing the existence of any integration/coordination policy in the countries studied, it focuses on the understanding of whether, when and how the need for professional, institutional and organizational integration/coordination came to be formulated. Our approach seeks to deconstruct the meaning of integration and coordination policies led in the three countries, and to understand their institutional, organisational and historical contexts. This paper presents the first results of a literature review and an empirical survey by semi-directive interviews led since 2013 in France (n=27), Sweden (n=22) and the UK (n=28). It covers two levels: interviews with national experts at the macro level of policies developed and implemented nationally and with actors at the micro level concerning specific schemes implemented locally, to take into account the organizational particularities in each of the countries and on chosen territories. Results: The analysis of the policies in France, Sweden and UK shows the impact of both the national and the territorial contexts, which varies from one country to the other. It leads us to identify three different policy-making processes towards coordination/integration. We provisionally named them “layers”, “chains” and “lineages”. These logics are unequally shared out between the three countries: the layering process can be identified as a dominant trend in France whereas the chaining logic is dominant in Sweden. The lineage metaphor, finally, looks more appropriate when it comes to analyze the history of attempts for reducing fragmentation in England. Yet, further analysis including concrete solutions experimented on both national and local levels shows that there come to be a mix of logics mobilized in each concrete situation. Thus, the different logics are not directly opposed and they can be expected to work together. Discussion: Qualitative comparison raises significant challenges due to the risk of comparing schemes or elements which are not comparable, particularly for locally implemented models developed in specific territorial contexts, unless one focus on the learning and adaptation processes. Secondly, comparison between countries representing different welfare models runs the risk of reducing the observed outcomes to the same variables used for the identification of the welfare models. By putting the accent on identifying the underpinning logics of integration/coordination policies in the three countries rather than running a systematic comparison of their content, we address these challenges in two ways: - The three logics identified articulate both macro and micro analysis since they are based on the identification of the mechanisms underpinning both national policies and locally implemented measures. -The three logics are analyzed in relation to the different welfare state models and to the wider political projects they belong to. Conclusion: The next step of the research consists in identifying common cross-cutting issues in at least two of the three countries studied: -The impact of the political variable on the orientation and the content of integration/coordination policies; -Interactions between local and national levels as opportunities and constraints for integration; -The issue of New Public Management in the integration/coordination process. Taken as basis for comparison, these issues can inform aspects of processes of integration/coordination that are often analyzed in one single national/local context.