Restoration theories teach us that views to nature and spending time there help us restore from attention fatigue and socio-emotional stress. Research has also indicated that we have a beneficially adaptive taste, in that our preference for restorative environments increases with our need for restoration (Hartig & Staats, 2006; Staats et al., 2010). Such an adaptive appetitive system would help us seek out more restorative contexts when we need them. In the current studies, we wanted to investigate this adaptive system further, exploring whether need for restoration would not only have us prefer natural elements more, but also direct our visual attention towards natural elements – opportunities for restorative escape – faster and more explicitly. We used eye tracking to investigate the elements individuals scan and fixate on when presented with new scenes.Eye movements are rapid and largely beyond our conscious control. In between movements the eyes briefly fixate on specific elements in a scene. This scene scanning involves both top-down and bottom-up processing. Bottom-up processing involves matching the position of certain critical objects with stored cognitive maps. Top-down processes were first demonstrated by Yarbus (1967), who showed that cognitive tasks will influence which objects are attended to. Whether restorative need impacts such top-down processes in scene processing is as yet unknown. Our research question therefore was: Does need for restoration influence scene processing; more specifically, do people fixate more, and more quickly, on natural elements of scenes after stress induction?We performed two experiments (N=60 and N=50), in which we fitted participants with an eye tracker, induced stress (vs. no stress in the control condition), and then presented them a set of pictures. In Study One we employed an emotional stressor: we induced a mood by asking participants to recall an episode in their life where they experienced this emotion (e.g., Hucklebridge, 2000). In Study Two we used a cognitive stressor: the Markus & Peters Arithmetic Test (MPATest, Peters et al., 1998). The images all contained mixed content (i.e. natural and man-made elements, which were coded as either natural or urban Areas of Interest (AoIs). In both studies we compared heatmaps (the duration of fixation on natural vs. urban AoIs) and scanpaths (the order in which participants scanned the image) between the stress and control conditions. The studies were ongoing at the time of submission; results will be available in December 2011. These will tell us whether our visual scanning system helps us seek out opportunities for restoration under stress as compared to neutral mental states.