Abstract
Uncertainty is almost unavoidable in environmental impact assessment (EIA) predictions, for complex and manifold reasons. In this paper, evidence is presented that decision-makers and other stakeholders are often not made aware that such uncertainty exists. Also, they are given only limited access to information about input data and the assumptions underlying predictions. It is argued that more emphasis should be given to improving the communication of uncertainty in EIA predictions and to making the prediction processes more transparent in order to improve EIA as a decision-aiding tool. The discussion is based on a study including 22 cases.