There is an increasing demand for transport decisions to be made in thelight of some form of what may generically be called cost benefit analysis(CBA). The actual techniques may vary and often deviate quite significantlyfrom the conventional economic concept of CBA but in all cases theprocedures tend to be information intensive. In some cases there is astatutory requirement that standardized parameters are used, but oftenspecific studies are performed. In these latter cases the studies arelengthy and generally financially costly. To reduce these problems therehas been an increasing advocacy of 'benefit transfers' being deployed. Thisinvolves taking parameters such as elasticities or valuations ofexternalities from one study and applying them to another. This paperexplores the validity of the benefit transfer approach from both atheoretical and an empirical perspective. In particular, it examines theextent to which it is legitimate to transfer parameters from one case andapply them to another and to isolate the conditions under which suchtransfers may sensibly be made.