A new editorial in Science questions the common understanding of “scientific consensus”, and I reflect on what makes science reliable.