The shift from institution-based mental healthcare system toward integrated care system is - not only but also- a question of organisation. Recognising the organisational dimension of this shift is a pre-requisite for improving the design of mental healthcare researches. In Belgium, the shift toward integrated care follows a change strategy inspired by the WHO: it is stimulated through pilot projects encouraging the transformation of mental healthcare delivery at the local level. Scientific evaluations of these projects evidenced that, despite the development of innovative practices, the shift is still not complete. The question of how innovative practices are arising and how to support their generalisation nevertheless remained overlooked. Theory: In this paper, we argue for reconsidering the design of scientific research on mental healthcare reforms, by integrating key learnings from organisational theories. Organisation theorists showed that talks are constitutive of organising process, particularly in complex environments where multiple interests, ideologies, and types of expertise are coexisting. The importance of talk in re-organising work processes, particularly inter-professional meetings, has been particularly emphasised in mental health services. We analysed the impact of inter-professional meetings on the reorganisation of mental healthcare in Belgium by relying on qualitative methods and the principle of triangulation: inter-professional meetings n=72 were observed as part of two scientific researches[1]; the meeting participants were interviewed n=57 and interview and observation data were confronted to documents linked to the observed meetings. Results: Based on excerpts of interviews and transcripts of observation, the paper evidences that: - meetings most of the time fail to reach their expected outcomes; - the meeting context and leadership style of the chair importantly influence their real outcomes; - despite the feeling that meetings are useless, they bring about changes in mental health professionals’ representations and ways of working. Discussions: Mental healthcare researches have underestimated the influence of inter-professional meetings on mental healthcare services’ transformation. Designing scientific researches integrating meeting as a research object to understand how the transformation occurs and instrument to experiment with different meeting techniques might help to address this gap. Conclusion: Two factors explaining the meeting contribution to innovation are: - the level of connectedness/separateness between the meeting context and institutional environment: increased separateness means increased possibilities to experiment with new roles and representations; - the management of meeting conversation: the meeting contribution to innovation is related to the level of ambiguity afforded by the conversation. Lessons: if meetings are supposed to bring solution to complex problems, they have first to be carefully prepared and managed, thus both researchers and practitioners might consider that meeting is a problem before being a solution. Limitations: the research was limited to the Belgian context and the results have not been generalised and tested in other countries and based on quantitative methods. Perspectives: Testing the results in other contexts and designing experimentation with meeting techniques for stimulating innovation in mental healthcare systems.