OBJECTIVES: Quantitative analysis of the physical and demographic parameters of access to Thames Chase Community Forest (TCCF), and how these have changed between 1990 and 2003; and qualitative exploration of our understanding of the links between health and the natural environment (TCCF), with a focus on the issue of 'access' to green space. STUDY DESIGN: Multimethod design involving both quantitative (analysis of physical access to green space) and qualitative (ethnography) components. METHODS: Quantitative analysis, using geographical information systems, of physical access to the community forest; and ethnographic research including participant observation, non-participant observation, in-depth interviews and attendance at meetings and conferences. RESULTS: The quantitative analysis showed that public access to green space improved between 1990 and 2003 as a result of the regeneration and acquisition of new areas, and the average reduction in distance to green space was 162 m. However, such improvements were distributed differentially between population groups. In both 1990 and 2003, people from deprived areas and in poorer health had better access to green space than people from less deprived areas, but the greatest improvement in access to green space over this interval occurred in areas of below average deprivation (i.e. in the more affluent areas). The ethnographic research showed different interpretations of the notion of access. Use of TCCF was determined by a variety of factors including whether a person could 'imagine themselves' using such a space, different perceptions of what is actually being accessed (e.g. a place to exercise or a place to socialise), and ideas about using the countryside 'properly'. CONCLUSIONS: The health benefits of using a green space, such as TCCF, for walking or exercising are well recognized. However, whether people choose to use local green space may be determined by a variety of factors. These are likely to include physical distance to access of green space, as well as perceptions and understandings of what is being accessed and how it should be used. This study has also illustrated the ways in which multiple methods can be integrated in public health research, and the merits of different approaches to undertaking multidisciplinary work of this type.