Many scientists feel that scientific outcomes are not sufficiently taken into account in policy-making. The research reported in this paper shows what happens with scientific information during such a process. In 2001 the Dutch Ministry of Transport, Public Works and Water Management commissioned their regional office in Limburg to assess how flood management objectives could be achieved in future in the Dutch Meuse valley, assuming climate change will increase peak discharges. To ensure political support, regional discussion rounds were to help assess the measures previously identified. This paper discusses the ways in which hydrological and hydraulic expertise was input, understood and used in this assessment process. Project participants as a group had no trouble contesting assumptions and outcomes. Nevertheless, water expertise was generally accepted as providing facts, once basic choices such as starting situation had been discussed and agreed. The technical constraints determined that politically unacceptable measures would have to be selected to achieve the legally binding flood management objective. As a result, no additional space will be set aside for future flood management beyond the already reserved floodplain. In this case, political arguments clearly prevail over policy objectives, with hydraulic expertise providing decisive arbitration between the two.