Is electoral sociology wrong to ignore, as it almost always does, the inequalities which find their way into the electorate from one election to another solely because of the differences in mortality and fertility between social classes? In order to answer this question, it is first necessary to clarify the phenomenon to be investigated. The construction of a theoretical diagram gives form, on the basis of fictitious data, to the way in which demographic factors affect electoral results, and makes it possible to determine what method of investigation to use (cohort analysis) and the nature of the demographic impact looked for. When an attempt is then made, using French data, to evaluate the effects of demographic inequalities on election results, it has to be concluded that neither differential mortality according to socio-occupational category, nor differential mortality according to sex, nor differential fertility have had significant effects in France. While electoral sociology would thus appear at present to be right, the fact remains that any research concerned with a population investigation over time could likewise give some thought to demographic changes which alter its subject and, in so doing, make use of cohort analysis.