A recurring finding in the landscape preference research is that there are differences in preference between different scene types. One characteristic of these experiments however was that the number of different types of scenes in any one experiment was small and there has been no systematic attempt to examine and account for the scene type effect. In a recent experiment we collected data relating to a number of different types of preference _ global preference judgments of the type that is general in the field and preference as a place to live and work or visit on a vacation _ together with judgments of whether the scene was natural or built. The data were collected using matched sets of a systematic selection of twelve scene types from two geographic locations with participants from each location making judgements of each scene type. Although our main interest was riot in differences between scene types as such, the results of the experiment demonstrated that there were very large differences between the scene types and that these effects were much larger than, and could not be accounted for by, the different models. The scene type effect was not only large but applied in a similar way to participants judging scenes from their home environment and scenes from outside of their home environment. The scene type effect would therefore appear to be an important facet of environmental preference. In examining our data however we became aware of a number of aspects which point towards an explanation for the scene type effect. The relationship between the different types of preference and the natural and built judgements suggest that the scene type effect may be related to another area of landscape research: the restorative effects of environments. The paper to be presented will review the results of our research from the perspective of this possibility and will report results from new experiments which obtained judgments on the restorative environment scale of the same set of scene types.