OBJECTIVES: A significant amount of literature documents the challenges of undertaking evaluative research on the public health impacts of interventions in the non-health sector. However, few studies have investigated why such studies are undertaken despite the undoubted challenges. Taking housing as a case study, the authors aimed to identify the factors contributing to successful evaluative research in the non-health sector. STUDY DESIGN: Qualitative interview study. METHODS: Semi-structured interviews with 16 investigators involved in seven successful experimental studies of housing interventions across the UK, analysed using thematic content analysis. RESULTS: Intervention studies were undertaken when existing collaborative links enabled 'windows of opportunity' to be exploited. Although different 'cultures of evidence' were reported across the collaborating teams, these did not necessarily map onto the public health research/non-academic divide, and did not undermine collaborative work when all parties could gain from taking part in the research. CONCLUSIONS: Focussing on success, rather than failure, suggests that to encourage the uptake of evaluative evidence in the non-health sector, efforts might be better directed at fostering opportunities for partnership building rather than simply on educating non-health partners in the principles of academic research.