Choice experiments are designed to account for variations in environmental resources and site characteristics, as well as potential implications of these variations for willingness to pay. This may render choice experiment results highly suitable for benefits transfer. It is unclear, however, whether the flexibility of choice experiments renders the similarity of study and transfer sites less critical for transfer validity. Drawing from identical choice experiments conducted in different Rhode Island communities, this model assesses the extent to which error in function-based benefits transfer is related to the similarity of communities across a variety of observable dimensions. Results suggest that site similarity, at least across some dimensions, influences the validity of choice experiment benefits transfers. However, the use of some measures of similarity as indicators of transfer error may provide misleading results.