The broader study is of 265 backyards and 330 backyarders (a number of couples were included) in Alice Springs, Sydney and Wollongong (Head and Muir 2006, 2007). Our sampling strategy was designed to encompass the socioeconomic and geographic variability in each of these areas (Commonwealth of Australia 2002). Participants were recruited through media advertisements and appeals, letterboxing, snowballing from other participants, and by liaising with community groups. Each backyard was visited and a semi-structured interview undertaken on site with the participant/s by one of a team of three researchers. Questions related to the activities of different members of the household, changes that had occurred over time, people’s feelings about the space, what sorts of plants and animals were considered to belong, wider environmental attitudes and practices, and major influences. None of the initial questions was explicitly about water, but water emerged consistently in conversations about a variety of topics. The backyard was mapped and photographed, and checklists on the demography of the household, the structures in the backyard and the biogeography were completed. The interviews were transcribed and imported into the qualitative data-analysis program, N6. Initially, all water comments were content-coded for the context in which they were talked about and the practices described. Using a discourse-analysis framework, we coded for different kinds of motivations and investments.