The buildings and physical spaces in schools are strongly regulated by external funding arrangements and design specifications that are seemingly beyond the sphere of influence of most of those who work, play and study in them. Yet an emerging literature on the impact of the built environment on learning and behaviour (see Higgins, Hall, Wall, Woolner & McCaughey, 2005; Victoria Institute of Teaching, 2008) provides the rationale for local action to change the physical conditions that affect student engagement and teacher pedagogy. The schools involved in this study had all made significant changes to their built environment. These included: Removing doors to promote greater student movement Designing new learning spaces that enabled students to work collaboratively in groups Replacing walls with glass panels to make classroom more open and public spaces Removing traditional classroom furniture (desks and chairs) and replacing it with café-style seating Installing Wi-Fi to enable students to collaborate on-line Designating common spaces for student gatherings and celebrations From a policy enactment perspective, these interventions into the previously tightly controlled area of classroom and building design were quite bold and challenging initiatives. While some school leaders used formal funding approval processes to achieve their building programs, others acted unilaterally in defiance of directives ‘from above'. For example, one school leadership team enlisted the students to construct new furniture (benches with cushions) and to paint classrooms in brighter colours. Another commissioned a private architect to design a new school hall that exceeded the size specifications set by ‘the system'. The school leader paid for the new building by diverting funds from other areas of the school budget. These entrepreneurial initiatives demonstrate the power of local school leaders to defy external policies and regulations to achieve the teaching and learning outcomes that they and their senior leadership teams valued. The focus of this paper is on these key actors and the ways in which they ‘work upon one another and themselves in ‘doing policy'' (Ball, et al., 2011, p. 611).