Two identical open-ended contingent valuation surveys assessing willingness-to-pay for better protection against flooding were administered in 2005 and 2010 at the same site. The 2010 survey was administered to the same respondents as those interviewed in 2005 as well as to new participants. This experimental design allows us to separate the pure temporal dimension from the spatial and social dimensions of transferability, thereby permitting an investigation of the temporal reliability associated with a transfer of value estimates over a 5-year time horizon. Having isolated the pure effect of time, the design further allows assessing the social dimension of transferability.