Many of the earliest studies that sought to test the validity of benefit transfer, which typically involved the use of contingent valuation or travel cost applications, produced negative findings. One of the recommendations from this early testing was to trial alternative methods to produce source studies for benefit transfer, including multi-attribute techniques. Subsequently, there have been a number of studies conducted that have sought to test the validity of benefit transfer using choice modelling. The purpose of this paper is to review these studies, and to assess the current state of knowledge regarding the validity of using choice modelling for the purpose of benefit transfer. We also describe and consider the merits of an alternative approach for extending the use of choice modelling in benefit transfer, Bayesian benefit transfer.