The Belgian government Integreo program (www.integreo.be) aims at a nationwide whole-system change in health and social care towards more Integrated Care (IC), by means of geographically defined integrated care projects (ICPs). Starting in early 2018, the twelve ICPs are implementing their action plan to change the organization of their regional network, covering up to 150.000 inhabitants, towards more IC. They receive limited financing: one FTE coordinator. As the final goal of the national plan is to anchor successful changes and scale-up best practices, FAITH.be (Federated consortium for Appraisal of Integrated care Teams in Health in Belgium) developed a mixed methods realist evaluation framework which aims at responding to the questions if, how, for whom and why the implementation of the ICP is successful or not so far, and learn lessons to help the ICPs to adjust their governance to support successful implementation of a whole-system change in their region. We will present preliminary insights of the implementation analysis, describing the main difficulties ICPs struggle with and how they address them. We use participative action and realist methods following RAMESES II standards (http://www.ramesesproject.org/). Data collection tools are chosen in consultation with stakeholders: coordinators, partners of and policy makers. Intermediate results are fed back and discussed with the stakeholders. Two methods are combined: qualitative enquiry for all ICPs and case study methods for three selected ICPs. Normalization Process Theory (NPT) is chosen to guide the analysis. First observations show that ICPs struggle with divergence in vision about priorities and timelines between ICPs, Governmental agencies at both federal and regional level. After postponement of crucial policy decisions and delivery of promised ICT-tools, ICPs are forced to adapt their action plan, which impedes their coherence. Moreover, due to the limited financing, ICPs are expected to pre-finance their actions towards IC, which impedes the ...