France may be the most nuclearized country in the world, support from French public opinion has gradually but largely eroded over the last decades. However, since 2020, the support has considerably risen to reach a rate 60% agreeing to use nuclear energy in the future. How can one explain such a shift in such a short time? Using longitudinal European and French opinions polls from 2020 to 2024 as well as web-data taken from the French-speaking X (ex-Twitter), we mobilize the RAS model (Zaller, 1992 - RAS standing for Receive-Accept-Sample, a model which assumes that the opinions expressed reflect the messages received and accepted that are most salient at the time of the survey, and that there are no real structured political belief systems) as well as notions of ‘spiral of silence’ (Noelle-Neumann, 1984), ‘focusing event’ (Birkland, 1997; Kingdon, 1995) or ‘latent public opinion’ (Warshaw, 2017). The French surveys mobilized notably have an experimental aspect, trying to seek how strongly individuals hold to their opinions when presented with contradictory arguments. Part of those changes on a European scale can be explained by the geopolitical context opened by the war in Ukraine – understood as a focusing event whose ramifications weigh equally on all countries. Hence a contextual effect that can be measured in the French and European polls on the general public opinion. But some aspects of this evolution clearly depend on national events such as a reframing of energy policies intertwined with economical or employment during political elections. Data from X here shows the impact of local elections and then the presidential one on the French energy debate, the arguments used and the reconfigurations of interested communities. Initially restricted to communities of experts and politicized people, an argumentative effect can be measured in our surveys for the general public since 2023 – the same arguments being more performative in 2024 than in 2020. Hence, those moments constitute successful tests for nuclear energy in a general context more favourable than ever. In the end, we aim at explaining the chaining of sequences that overlap on local, national and European scales, with impacts the general public, and at understanding how public opinions may change so sharply and quickly even though technologies and infrastructures have not changed over this temporal sequence.