Just as temperature is measured in degrees and distance in metres, it has been accepted that a population’s evolution can be assessed by its ageing. Each month for the past twenty years or so, the press has produced a batch of articles on how France’s future is being jeopardised by its ageing citizens: its pension funds will soon be empty, its economic competitiveness and general dynamism weakened, and costs for health and social services will soar. No one would contest, however, that the contemporary sexagenarian has little resemblance to counterparts living between the wars and still less to ancestors from the early nineteenth century. The sexagenarian’s expectations of enjoying good health in the years ahead have risen considerably, and his or her place in successive generations has become central. Is there not a glaring contradiction between these changes in age’s reality and the notion of an ‘ageing’ population? It is the notion’s scientific pertinence, which for half a century has been built on unchanging statistical categories, that I should like to examine here, as well as the principal consequences of its utilisation.1