The assumption that improving housing conditions can lead to improved health may seem a self-evident hypothesis. Yet evidence from intervention studies suggests small or unclear health improvements, indicating that further thought is required to refine this hypothesis. Articulation of a theory can help avoid a black box approach to research and practice and has been advocated as especially valuable for those evaluating complex social interventions like housing. This paper presents a preliminary theory of housing improvement and health based on a systematic review conducted by the authors. Following extraction of health outcomes, data on all socio-economic impacts were extracted by two independent reviewers from both qualitative and quantitative studies. Health and socio-economic outcome data from the better quality studies (n = 23/34) were mapped onto a one page logic models by two independent reviewers and a final model reflecting reviewer agreement was prepared. Where there was supporting evidence of links between outcomes these were indicated in the model. Two models of specific improvements (warmth & energy efficiency; and housing led renewal), and a final overall model were prepared. The models provide a visual map of the best available evidence on the health and socio-economic impacts of housing improvement. The use of a logic model design helps to elucidate the possible pathways between housing improvement and health and as such might be described as an empirically based theory. Changes in housing factors were linked to changes in socio-economic determinants of health. This points to the potential for longer term health impacts which could not be detected within the lifespan of the evaluations. The developed theories are limited by the available data and need to be tested and refined. However, in addition to providing one page summaries for evidence users, the theory may usefully inform future research on housing and health.