The Built Environment Restoration Support Scale (BERS) was developed to measure perceived restorative potential of environmental components. The scale was based on scores from statements regarding perceived restoration in 14 architectural details. These details corresponded to 56 items, inquiring on restorative factors like being away, extent, fascination, and compatibility. The aim of this study was to test how well the BERS statements, or measurement items, fit to a unidimensional concept of restoration potential in an experimental setting by using Rasch analysis. The study objectives were to determine the internal scale validity and reliability of the BERS. The basic assertions of the Many-Faceted Rasch (MFR) model of the BERS were: (1) the more restoration supportive an item is, the more likely it is to be rated higher on the scale and (2) the more perceptive a participant is to restoration, the more likely that he or she will rate higher on the items as compared to participants who are less perceptive to restoration. The results show that, altogether, seven items which did not fit the MFR model of the BERS were removed. These items were: View Compatibility, Color Compatibility, View Fascination, View Being away, Plant Fascination, Plant Compatibility, and Plant Being away. The remaining 49 items, which had MnSq • 1.4 and z