Decision making for climate change management seldom incorporates uncertainty in the analysis that underpins the policy process. First, uncertainty is seldom characterised fully, and attempts to reduce uncertainty—when this is possible—are rare. Second, scientists are ill-equipped to communicate about uncertainty with policy makers, and policy makers most often favour pretended certainty over nuance and detail. Third, the uncertainty analysis that may have been conducted most often fails to actually influence policy in a significant manner. The case is made for (i) characterising and, to the extent possible, reducing uncertainty, (ii) communicating uncertainty, and (iii) reflecting uncertainty in the design of policy initiatives for climate change management. Possible elements for a research agenda on each of these areas are proposed.