Analysis of uncertainty is often neglected in the evaluation of complex systems (such as predictive models in hydrology or ecology, or environmental processes). Decisions and actions based on such systems may be error-prone for a variety of reasons, including lack of information, input errors or data variability. Research in uncertainty addresses this problem by investigating the causes of uncertainty and the characterisation of uncertainty associated with limited information. The article describes and compares definitions and terminology in uncertainty analysis and reviews suggested classifications. Traditional risk assessment is not directly equivalent to uncertainty analysis. The authors discuss distinctions and applications in various contexts relating to environmental management.