There are many definitions of ignorance in the context of risk assessment and risk management. Most refer to situations in which there are lack of knowledge, poor basis for probability assignments and possible outcomes not (fully) known. The purpose of this paper is to discuss the ignorance concept in this setting. Based on a set of risk and uncertainty features, we establish conceptual structures characterising the level of ignorance. These features include the definition of chances (relative frequency-interpreted probabilities) and the existence of scientific uncertainties. Based on these structures, we suggest a definition of ignorance linked to scientific uncertainties, i.e. the lack of understanding of how consequences of the activity are influenced by the underlying factors. In this way, ignorance can be viewed as a condition for applying the precautionary principle. The discussion is also linked to the use and boundaries of risk assessments in the case of large uncertainties, and the methods for classifying risk and uncertainty problems. © 2010 Elsevier Ltd. All rights reserved.