Natura 2000, the nature network based on the European Bird and Habitat Directives, is explicitly grounded on ecological science. To acquire a permit under the Dutch Nature Conservation Act, an appropriate assessment of significant effects must be conducted based on the best available scientific knowledge. In this way the scientific and policy world are directly linked. This article focuses on ‘significant effect’ as a boundary object to analyse how science–policy interactions shape the meaning and assessment of significant effect and how these interpretations influence the decision-making process. To this end, two conflicts over significant effect are investigated: the conflict over the 2006-spring permit for the mussel seed fishery, and the 2011 permit for the planned World Championship powerboat races. In both cases nature organisations started a court process against the government-granted permits in protest to the ‘‘no significant effect’’ claim, stating that there was insufficient certainty for this conclusion. These conflicts are approached as controversies between discourse coalitions with different interpretations of the ecological knowledge. We show how significant effect became a focal point in the controversies, limiting the debate to ecological arguments and science-based expertise, but also creating options for parties to advance their protest by articulating uncertainties. Only uncertainty of incomplete knowledge was explicitly addressed, excluding ambiguity of values and unpredictability of the actual ecosystem. We suggest that acknowledging the value aspect in disputes on significant effect would leave more space for effective solutions of the problems under debate.