Background: It is rare that decisions about investing in public health interventions in a city, town or other location can be informed by research generated in that specific place. It is therefore necessary to base decisions on evidence generated elsewhere and to make inferences about the extent to which this evidence is generalisable to the place of interest. In this paper we discuss the issues involved in making such inferences, using physical activity as an example. We discuss the ways in which elements of the structural, physical, social and/or cultural environment (environmental factors [EFs]) can shape physical activity (PA) and also how EFs may influence the effectiveness of interventions that aim to promote PA. We then highlight the ways in which EFs may impact on the generalisability of different types of evidence. Discussion: We present a framework for thinking about the influence of EFs when assessing the generalisability of evidence from the location in which the evidence was generated (place A) to the location to which the evidence is to be applied (place B). The framework relates to similarities and differences between place A and place B with respect to: a) the distributions of EFs; b) the causal pathways through which EFs or interventions are thought to exert their effect on PA and c) the ways in which EFs interact with each other. We suggest, using examples, how this scheme can be used by public health professionals who are designing, executing, reporting and synthesising research on PA; or designing/implementing interventions. Summary: Our analysis and scheme, although developed for physical activity, may potentially be adapted and applied to other evidence and interventions which are likely to be sensitive to influence by elements of the structural, physical, social and/or cultural environment such as the epidemiology of obesity and healthy weight promotion.