Country parks provide visitors with an excellent environment that can facilitate good health by helping relieve these stress-related symptoms that accompany modern life (The Countryside Agency, 2004). Health promotion strategies that incorporate a natural element may prove valuable to an increasingly urbanized population (Hartig et al, 2003). Restoration is the process of returning a person back to their original healthy condition from previous physical and/or psychological deprivation (Han, 2003). On the whole, natural environments appear to better serve physiological, emotional and attentional restoration than urban environments (Berto, 2005). But natural environments contain many sources of potential danger that may evoke negative reactions, including natural threats such as predators, animals and lightning (Tooby & Cosmides, 1990) but also the threat of being attacked by another (Coble et al, 2003). Indeed enclosed, dark and dense wooded areas can prove intimidating rather than therapeutic (Milligan & Bingley, 2007). If this proposition is correct, natural environments that are perceived as dangerous may face dwindling visitor numbers and the possibility of losing some of their restorative potential, reducing the effectiveness of health promotion strategies that they may incorporate. The relationship between perceptions of danger and anticipated restoration has yet to be examined. Using Fisher and Nasar’s (1992) typology of safety that utilises Appleton’s (1975) Prospect Refuge model as a framework, the first part of this research attempted to explore the relationship between perceived danger and anticipated restoration within a country park environment before secondly, examining the practical use of the framework as a potential tool for advising environmental design so as to maintain visitor numbers and to secure the restorative benefits of contact with such environments by minimizing perceptions of perceived danger. Respondents were randomly assigned to one of three conditions based on Nasar and Fisher’s (1992) model. The conditions used photographs taken within a local country park that had been pre-tested as having either low, medium or high levels of prospect; refuge for potential threats and opportunities for escape. Participants then viewed a series of slides depicting this walk before being asked to complete items measuring perceived danger and anticipated restoration. Anticipated restoration (as measured by Han’s SRRS) was found to be strongly negatively related to perceived danger, with anticipated physiological and emotional restoration being the strongest negative predictors of perceived danger. Scenes with low levels of prospect and accessibility but a high number of hiding places were perceived as more dangerous and less restorative than scenes high in prospect, accessibility with few hiding places. These results support both ART (Kaplan, 1995) and SRT (Ulrich, 1983). They also suggest that the design of visitor friendly natural environments such as country parks need to minimize perceptions of danger to ensure the anticipated restorative value of the environment is not degraded. One way of doing this appears to be through the use of Nasar and Fisher’s (1992) by manipulating prospect, refuge and escape. The next stage of the research is to examine the effect of specific types of danger on anticipated restoration. Preliminary analysis has revealed that social danger (the threat of being attacked by somebody) has a greater detrimental effect on anticipated restoration than either physical danger (the threat of an accident such as tripping or being attacked by an animal) or the threat of becoming lost. However, the use of Fisher and Nasar’s (1992) model does not seem to work as no significant differences in social danger were found between the 3 conditions. The results also found that some people anticipate some sort of physical danger to be restorative. Although physical danger can be anticipated as restorative to some, social danger appears not to be, and so such environments must find a way of minimising perceptions of social danger. People may stereotype social danger in natural environments, so I hope to examine the effect of prior knowledge/media stories of social dangers within natural environments to see whether anticipated restoration is affected when people are presented with differing degrees of social danger that are either related or unrelated. I also hope to conduct some qualitative field studies combined with some objective measures of restoration to help explore the perceived dangers associated with country parks, their actual restorative effects and to devise effective strategies to minimize social danger. As perceived danger does Paper in Young Researchers’ Workshop appear to play a role in anticipated restoration, I also hope to use the results of these studies to try and improve the practical utility of existing measures in design decisions by integrating an element of perceived safety/danger.