In the analysis of mortality statistics high age-specific death rates could be interpreted as meaning more deaths (more disease), but they could equally well be interpreted as meaning earlier deaths (death at younger age). The distinction markedly affects the choice of hypotheses that may be advanced to explain variations in person, time and place and the design of subsequent, more detailed field studies to test the hypotheses. Furthermore, the majority of descriptive papers make no comparisons with a control disease and thereby break one of the ground rules of epidemiology. This paper shows how, in the example of the geographical variations within England and Wales of ischaemic heart disease, a control may be simply introduced and that much of the observed variation is not in the proportion who suffer heart disease deaths but is in the age at which deaths occur.