Significant associations have been found by a number of researchers exploring the relationship between green space and stress. People’s preferences for stress regulation have also been shown to vary by activity and the environmental characteristics of green space (Stigsdotter and Grahn, in press). This current study builds on an earlier pilot indicating that people living in areas of high urban deprivation in Scotland may experience stress differently according to the levels of green space in their living environment (Ward Thompson et al, under review). Our aim was to explore, in a similar population, preferences for stress regulation in relation to self-reported stress and wellbeing and a range of environmental attributes. The research design grew from focus groups carried out with residents in four case study areas in Central Scotland characterised by high levels of deprivation and high versus low levels of green space (n= 29). Findings indicated that preferences for stress regulation appeared to vary in relation to the percentage of green space in the living environment. To explore this further, we used adaptive conjoint analysis, coupled with hierarchical Bayes estimation, to assess activity preferences when choosing a place for stress regulation. Via a computer aided household survey, adult residents living in areas of high and low green space in Edinburgh and Dundee, Scotland (n=406) were presented with a series of scenarios in relation to a question on the need to ‘get away’ when under stress: ‘where would you go to clear the head?’ Participants selected one of four behavioural options based on findings from the focus groups; to seek peace and quiet, to go for a walk, to seek company, or to stay at home. For each option, attributes of the environment were offered in combinations, to explore preferences for achieving stress relief. The attributes – selected on the basis of focus group findings – were specific to each chosen behavioural option but included location, activity, traffic density, distance to preferred location, views, greenspace and the social context. Self-reported stress and wellbeing measures were also captured.Conjoint analysis was run to prioritise the attributes that were most important for achieving the chosen behaviour. Early analysis for all respondents shows location, activity and views to be important attributes in relation the chosen behaviours, but that this varies by sub-groups, for example, by gender and employment status. The full findings will be presented and discussed in relation to the potential of green space to promote stress regulation.