As a historiographical category, ‘big science’ was elaborated from the point of view of advanced countries. However, some developing countries decided to invest a significant part of their rather modest science budgets in building many-million-dollar facilities. A comparative approach to the study of the first stages of the Argentine TANDAR heavy ion accelerator and the Brazilian National Laboratory Synchrotron Light (LNLS) projects may help understand specificities in patterns of organisation of big science in peripheral contexts. Oversimplification of the decision-making processes linked to authoritarian political contexts—which allowed to overcome the lack of consensus within the physics community as well as financial uncertainties—seem to have been a necessary condition for TANDAR and LNLS, which differentiated them from big science in developed countries. Additionally, the different ways in which the institutionalisation of the nuclear area took place in Argentina and Brazil seem to have been responsible for the different paths followed by experimental physics between the 1960s and 1980s: nuclear physics in Argentina and particle physics in Brazil.