Many studies on environmental preference have indicated that people prefer “natural” scenes to “human-made” scenes (e.g., Herzog, 1989; Purcell, Lamb, Peron, & Falchero, 1994). Researchers argue that strong preference for natural scenes is associated with biological importance of natural elements for people or restorative quality people feel in natural environments (e.g., Kaplan, 1987; Purcell, Peron, & Berto, 2001). Since a significant number of studies have shown the similar result, it appears that preference for natural scenes has been taken for granted. However, a study by Nassauer (1993) provided interesting findings that may question the commonly accepted association between preference and naturalness. Comparing people’s perceptions of conventional lawns and various patterns of native plant gardens in a residential area in the US, Nassauer found that the perception of tidiness and care is highly correlated with preference. Her study suggested that tidiness is another salient dimension that affects preference. However, since the stimuli in her study were almost equal in terms of naturalness, Nassauer did not touch on the relation between naturalness and tidiness.In order to closely examine the role of naturalness and tidiness, this study investigated people’s perception of various types of outdoor scenes, which range from tidy to messy and from natural to human-made. The objective of the study was to explore in what way the perceptions of naturalness and tidiness interact and influence preference for the environment. One hundred and seventy-eight undergraduate students in Japan and Australia assessed eleven photographs of outdoor scenes on a semantic differential scale. The scale involved ten pairs of adjectives including liking, attractiveness, beauty, interestingness naturalness, simpleness, conspicuousness, unclutteredness, maintenance and efficiency.Factor analysis (oblique rotation) on the ten semantic differential items extracted three factors. The first factor included liking, attractiveness, beauty and interestingness. This factor can be interpreted as preference. The second factor included naturalness (negative), simpleness (negative) and conspicuousness. If these items are interpreted in a reverse way, it is possible to characterise this factor as plainness. The third consisted of unclutteredness, maintenance and efficiency. The factor involves neatness and good human care, thus can be interpreted as tidiness. The results indicated that preference and plainness were fairly independent from each other, but preference and tidiness were somewhat correlated. The perception of naturalness was mainly associated with plainness, but it had a moderate loading on preference as well. Bivariate correlations between liking, naturalness and unclutteredness were also calculated. A significant but modest negative correlation was found between naturalness and unclutteredness. However, the similar magnitude of positive correlation was also found between preference and naturalness, and between preference and unclutteredness.The analysis suggests that salient perceptual characteristics of the presented stimuli are preference, plainness and tidiness. Naturalness did not appear as a primary perceptual dimension. However, it seems fair to say that naturalness has a complex role in the perception of the stimuli employed in the study. The results indicate that perception of naturalness has mixed contributions to preference: a positive contribution from natural elements (tree, foliage and water) and their restorative effects, and negative contribution from its rather untidy appearance. Thus, it can be argued that naturalness has two different aspects that have opposing influences on preference. In other words, naturalness is not a unidimensional construct, but involves at least two dimensions. Further research examining the dimensionality of naturalness seems worthwhile to have a richer understanding about the way people perceive natural environments. The results of the study will contribute to the design and management of various types of outdoor environments. Past research has demonstrated that ecological design is often perceived as being messy and unkempt (Mozingo, 1997). This study will provide a theoretical underpinning for design and management principles of such environments.