Following the theoretical approach of Herbold (1995), Gross and Krohn (2005), and Van de Poel et al. (2017), this article argues that nuclear waste management is a real-world experiment. Based on this first assumption, we examine how radioactive waste management (RWM) organizations conceive or organize their experiments. Through three illustrative case studies in France, Belgium and Canada, we highlight how the RWM organizations obliged to participate in complex networks and unable to completely control the experimental process, adopt two different attitudes: an “open” or “closed” experimental mindset. We argue that these mindsets provide different answers to the questions: which main variables to focus on, how and who should design them, how to deal with conflicts and unexpected events, what are the justifications for participation and expert analysis, and what are the expected outputs and outcomes. The findings underline that although some RWM organizations have -at least since the participatory turn- had some ‘open’ mindset moments in some cases, they quickly revert to a closed mindset. We conclude by emphasizing the need for practitioners and scholars to further examine and evaluate the virtues of the open mindset when the experimenter assumes the program has a real-world experimental status. This status recognizes the limits of control over experimental conditions, allows for more substantial moral considerations when making technical choices before wider audiences and allows for collective sharing of responsibility, knowledge production and trade-offs over such a long-term and controversial program.