One hundred and sixty-two women were evaluated for psychiatric morbidity in the second trimester of pregnancy and also six to eight weeks postnatally. A sizeable proportion of the women also had a second prenatal assessment in their third trimester and in the lying-in ward shortly after giving birth. Assessments were conducted with the 28-item GHQ. This was initially validated using a subsample of 106 women, taken from the original group, who were interviewed with the PAS. Thirty per cent of the women were ‘cases' at the first prenatal assessment, while only 14% were ‘cases' six to eight weeks postnatally. Thus, even though there was an overlap between prenatal and postnatal morbidity, there was also substantial difference between the groups that were symptomatic at both periods. This observation was reinforced by the low correlation between the mean GHQ scores at both periods. Both prenatal and postnatal morbidities were associated with recent adverse life events, with the latter more likely to be associated with marital and family events. This observation is in support of the view that neurotic problems, prenatally or postnatally, are caused mainly by psychosocial factors.