Over the past four years new research has been conducted on mental schemata mobilized by the common observer when s/he is engaged in visual perception of buildings, and on the surprising metamorphoses that subsequently occur, at the phenomenal level, even to the building representations. This line of research follows the studies presented in Berlin (lAPS 8, P. Bonaiulo, M. Miceu Romano & F. Bonaiuto, 1984) and Delft (lAPS 10, P. Bonaiuto, A.M. Giannini & M. Bonaiuto, 1988). We utilized three different experimental situations: 1. Actual buidings, viewed in situ. 2. Tridimensional models of these or other buildings, proportionally reduced, and tridimensional control objects. 3. Pictorial representations of buildings, and control pictures, with varying degrees of realism. The use of more or less intense contradictions of the expectations related to buildings - i. e. the reference to, and the utilization of, architectural and urban planning anomalies, was confirmed as a fruitful method of investigation. The focal architectural anomalies include well known anomalous buildings such as the Tower of Pisa or other touristically attractive places; and, more in detail: leaning buildings, structures with acute or obtuse corners, buildings with fractured, misaligned, concave or convex facades, distorted rooms, etc. Our general research paradigm allows the comprehension and prediction of the actual metamorphogenetic process of visual images corresponding to these incongruous buidings. The findings suggest that transformations may occur from a very regular or relatively regular image (attenuation of the actual anomaly), to a realistic image (anomalies are exactly appraized) or to a more or less emphasized image ( anomalies are over-stated); and vice versa. The transformations listed above are found as functions of the following independent variables: 1. Development, accuracy and strenght of the general mental schema of building; and, in addition, the properties of several sub-schemata (i. e., specific to public or private building, etc.). 2. Observation conditions: a) Ambiguous, when subjects receive less crucial perceptual informations, thereby enhancing the number of possible perceptual solutions. In these conditions subjects are more likely to attenuate the perception of anomalies, and assimilate the image to the mental schema. b) Intermediate, when subjects receive balanced informations favouring realistic perceptual solutions. C) Unambiguous, when subjects receive many and univocal informations, with an obliged perceptual solution, very different from the expectations. These subjects tend to exaggerate the perception of anomalies; promoting a contrast process between the image and the pie-existing mental schema. 3. Verbal or non verbal messages given to the observer. These are the oral definitions and descriptions accompanying the building presentation. Also changes in physiognomic qualities connected with the manipulation of colour, shape, texture or meaningful details (flowers, etc.), are very effective in producing metamorphoses of the building perceptions. 4. Observer's cognitive style. The way the subject elaborates the incongruous images is linked to personality trait constellations and mantain itself costant when facing different kinds of anomalies. The presentation of a leaning building model and the study of metamorphoses of its perceptual image related to the observation conditions is the main ingredient of a new very quick and promising psycho-diagnostic technique.