In spite of an extensive literature, which was reviewed in two previous papers (Gibson, 1954(Gibson, , 1956, there is still some difference of opinion on whether permanent and significant hypertension is a frequent sequel to pregnancy toxaemia in women who enter pregnancy with normal blood pressures; and again about the amount of damage which may be done by a toxaemic pregnancy in a woman already hypertensive. The three major difficulties of assessment are: (1) the difficulty of definition and classification; (2) the fact that few women have had their blood pressures taken before pregnancy started, and it is well known that pressures taken during the second trimester of pregnancy may give unusually low readings; and (3) the fact that, in a normal population, blood pressure varies with age, so that in a follow-up study final blood pressures must be expected to be somewhat higher than initial readings taken some years previously. For all these reasons one of us was about to embark on yet another follow-up study of a. small series of cases of pregnancy toxaemia observed a few years ago (Platt, Stewart, and Emery, 1958), but on reading the literature it seemed that the very much larger series of cases observed by Gibson (1954Gibson ( , 1956 (which is referred to in this paper as the Belfast series) was capable of further analysis. This paper is therefore a re-examination, jointly by physician and obstetrician, of the Belfast series.In this series Gibson took the figure of 140/90 to be the arbitrary dividing line between normal blood pressure and hypertension. If either the systolic or diastolic pressure was below this level the patient was regarded as normotensive. Of 1,420 toxaemic pregnancies in the series, 95% of the women were re-examined an average of four years later. Of these, 63.6% were adjudged, after review, to be cases of preeclamptic or eclamptic toxaemia, and 11.7% to be cases of chronic hypertension complicated by pregnancy. Toxaemic Series The main objects of the re-examination were to compare the initial and later blood pressures of these women with those of a general population of women in the same age groups. For our controls we have taken the population of 1,204 unselected females recorded by Hamilton, Pickering, Fraser Roberts, and Sowry (1954a), For the sake of brevity this is referred to as the London series. It must be noted, however, that our toxaemic series is not strictly comparable with a series of unselected women, because we have arbitrarily excluded anyone with a blood pressure of 140/90 or over, and the London series clearly shows that in the older age groups something approaching a third of all women are likely to have blood pressures higher than this. Table I and Fig. 1 show the systolic pressures and Table II and Fig. 2 the diastolic pressures of the toxaemic series in early pregnancy and again some four years after the pregnancy, set out as percentages and compared with the London series. It will be seen at a glance that the distribution of blood pressures of these women in early p...