Between the 1970s and the mid‐1990s the body mass index (BMI) became the standard means of assessing obesity both in populations and in individuals, replacing previously diverse and contested definitions of excess body weight. This article draws on theoretical approaches from the sociology of standards and science and technology studies to describe the development of this important new standard and the ways in which its adoption facilitated the development of obesity science, that is, knowledge about the causes, health effects and treatments of excess body weight. Using an analysis of policy and healthcare literatures, I argue that the adoption of the BMI, along with associated standard cut‐off points defining overweight and obesity, was crucial in the framing of obesity as an epidemic. This is because, I suggest, these measures enabled, firstly, the creation of large data sets tracking population‐level changes in average body weight, and, secondly, the construction of visual representations of these changes. The production of these two new techniques of representation made it possible for researchers in this field, and others such as policymakers, to argue credibly that obesity should be described as an epidemic.