Dealing with uncertainty has become a matter of great concern for policy makers and scientific research in a world facing global, epochal and complex changes. But in essence, you cannot entirely predict the future. This article aims at conceptualizing the limits to anticipate the future – or what is often referred as the substitution of risk for uncertainty. In contrast to most theories examining risk and uncertainty, we start from the assumption that there are limits in the substitution of risk for uncertainty and that distinguishing between ontological and epistemic levels of analysis helps clarify such limits. The paper makes two arguments: first, most approaches see no ontological and/or epistemic limit in the substitution of risk for uncertainty; second, the pluralization of science is the only way to cope with limits in substituting risk for uncertainty. This second argument draws on the assumption that accounting for the uncertainty of the future depends on knowledge production processes able to overcome disciplinary boundaries and better include lay and expert knowledge. In times of great concerns regarding mitigation and adaptation to the ecological crisis, we illustrate our arguments with insights from global environmental governance.