Silva et al. (Land Use Policy, 21 July 2020) offer an assessment of the links between deforestation, livestock production and exports in Brazil. Their analysis, based on relative changes in beef production and pasture area across the whole of Brazil, showed an “apparent decoupling of the link between beef production and deforestation in Brazil”. In reanalysing these links, we find that Silva et al. underestimate the strong, positive and significant associations between Brazilian livestock production and deforestation. Moreover, despite focusing the title, abstract and the beginning of their manuscript on the Amazon, their analyses are conducted at the national level, and fail to recognise marked differences in the development trajectories of Brazilian biomes, and that most of the recent pasture expansion in Brazil has replaced Amazonian forests. To progress any debate and aid decision-making regarding land-use changes in the Amazon, a region often in the spotlight and subjected to many debates that lack evidence, scientists must be open and scrupulous with their data sources and analyses. © 2020