The tension between how the concept of inclusive education is verbalised in various educational policies and practices, and how it is realised in classroom practices, is examined in this paper. The data derive from an ongoing research and development programme (Inkluderande lärmiljöer/Inclusive learning environments) in seven municipalities in Sweden, orchestrated by a research institute (Ifous) over three years (2017-2019). The aim is to enhance inclusion within the schools taking part in the programme, through for example lectures and seminaries arranged by the institute. The programme includes administrative management at municipality level, school management, teachers and student health professionals from across the school sector. It also includes a group of researchers, conducting different studies, with ‘inclusive learning environment’ as a common and overall orientation. This paper is theoretically framed by the ‘theory of practice architectures’ and ‘ecologies of practices’ (Kemmis et al, 2014). According to the theory, practices are shaped by three kinds of overlapping arrangements: Cultural-discursive arrangements such as discourses and languages affect what is possible to say in and about practice (e.g., deficit discourses, critical discourses, discipline-specific discourses, languages). Material economic arrangements – for example material, technological, financial, organisational, and other resources – affect what it is possible to do in practice (e.g., buildings, schedules, workload calculators, funding). Social-political arrangements are those that affect the ways in which it is possible for people to relate to others (and things and places) in practice (e.g., organisational rules, mandates, solidarities, hierarchies). These arrangements form the practice architectures of practices such as leading, teaching, and learning. The first analysis shows a gap between the different levels involved. The administrative management and the teachers have different views on whether or not there are student groups divided into smaller ‘special need’ groups. Furthermore, there is an ambition at the administrative level to transfer students, labeled as ‘students with special needs’, into ‘ordinary classes’. This ambition clashes sometimes with the teachers’ understanding of how to promote inclusion. The data also reveal a worry concerning ‘transitions’. Participants in the programme describe transitions, as ‘critical moments’ as well as ‘challenging’. The transitions include for example, the transfer of students from third grade to fourth grade or merging two classes together, including special education classes. How well these transitions are handled varies significantly amongst the schools in the programme.