The economic literature has devoted relatively strong attention to eco-labelling schemes. Nevertheless, while succeeding in some markets, they often fall short of their promises. We analyse the gap between the academic design of eco-labelling schemes and their real implementation. We contend that providing information is not enough. We then use recent advances in behavioural economics to inform policy makers on the potential of behavioural interventions in order to design better eco-labelling schemes. Policy relevance Many public policies, including eco-labelling schemes, are still based on an inaccurate description of human decision making, mainly borrowed from standard economics. However, numerous psychological and behavioural studies show that people reg-ularly behave in ways that contradict some standard assumptions of economic analysis. Departing from the conventional view that information-based policies such as eco-labelling schemes will quasi-automatically help mitigate issues such as climate change by guiding consumers' and firms decisions, we argue that information provision is necessary but not sufficient. Admitting that consumers' decisions are guided by factors other than price and information, and, taking systematically into account be-havioural biases can offer to policy makers low-cost levers with first-order effects in order to increase the environmental per-formance of eco-labels, or at least decrease the likelihood of counterproductive effects.