In a controlled trial pregnant women who were hypertensive before the 28th week of gestation were randomly allocated to treatment with methyldopa or no anti-hypertensive treatment. The children from these pregnancies have been re-examined at four years of age and their development compared with a random sample from the same maternity hospital population. Their health, height, weight, and the incidence of sight, hearing and speech problems did not differ. None had gross neurological abnormalities. Boys in the treated hypertensive group had significantly smaller heads than in the other two groups, but there was no correlation between head circumference and developmental score in this group (r = 0.020). On average the children in the random sample were the most advanced when assessed by a global score of development. In each developmental sector the mean score for the treated hypertensive group was consistently higher than the untreated hypertensive group. We conclude that maternal hypertension is associated with slight developmental delay in early childhood. There are some indications that treatment with methyldopa may reduce this effect.