Recovery practices following the loss of home, sense of security, space and possessions, have recently become a focus of UK government attention. How people recover from disasters is seen to have a direct bearing on individual, community and economic well-being, so that the recovery itself becomes a form of social change. A plethora of instruments: templates, checklists and guidance documents have been produced to effect this recovery. We term these ‘technologies of recovery’, which work within a wider context of disaster planning aimed at bringing order where much is uncertain, reactive and dependent on emerging relations between people, things and spaces. While such protocols are not necessarily unwelcome, they carry many assumptions. We show how these technologies are built from official, distal narratives, versions of recovery remote from situated practices or recovery-in-place. Official emergency planning builds on ‘lessons’ from previous emergencies, to be then applied to future crises. Knowledge that is situated, complex, and partial is potentially useless because emergency planners seek accounts that don’t depend on highly localised circumstances. From a five-year ethnography of both a flooded community and the development of government recovery guidance, it became clear that technologies or recovery became transformed and re-made in localised practice when enacted by newly formed and precarious collaborations of residents and local responders. Operating alongside, and sometimes underneath the official response, residents and local responders demonstrated a remaking of the politics of recovery.