The comparison of trends in ischaemic heart disease (IHD) mortality in different social classes is confounded by historical changes in diagnostic techniques and statistical classification, and possibly by different standards of diagnostic accuracy in the different social classes. The problems can be circumvented by taking advantage of the fact that in middle age (45 to 64) the IHD death rate is much higher in men than in women. This large sex differential is not present in any of the other causes of death with which IHD might easily be confused and it is therefore relatively unaffected by diagnostic errors and variation in classification. The changes that have occurred in the sex differential in Social Classes I and V in England and Wales between 1931 and 1971 confirm anecdotal clinical reports that the male vulnerability to IHD appeared first in Social Class I (professionsl). By 1971 Social Class V (unskilled) had caught up, and the men in these two social classes now experience an almost identical excess in cardiovascular death rate compared with their wives.