The use of reconstitution data for the estimation of adult mortality has always posed greater problems than its use for the estimation of infant and child mortality. The first problem in this connection does not appear capable of resolution. The parish registers do not yield any information which can be used to establish the presence of unmarried adults in the parish. The deaths of bachelors and spinsters may be registered as punctiliously as those of husbands and wives, but, if there is no way of establishing their presence in the parish except at the time of their death, no dependable mortality rate can be calculated. Flow can be measured but stock cannot. Any adult mortality rate, therefore, refers only to the married population. Since most men and women did marry, a rate which refers to the married population is likely to be broadly representative of the population as a whole, but, in considering the adult mortality rates produced from reconstitution data, this limitation should always be borne in mind.

The measurement of the mortality of married adults, however, also involves difficulties. Anglican registers very rarely provide information about age at marriage before Hardwicke's Act, and even after that act came into force the ages of the bride and groom were seldom given. If any information was given it might often take the form of stating that the bride or groom was over 21, or ‘a minor’, or ‘of full age’.