In 1994 we set up a specialist clinic for pregnancy dermatoses, both to improve the management of pregnant women with skin problems and to enhance our general understanding of the pregnancy dermatoses. This clinic has provided a large database of 200 women which has formed the basis for a prospective study over a 2-year period. In each case the dermatological diagnosis was clearly defined on clinical criteria, with additional help from histopathology and direct immunofluorescence of the skin where appropriate. We have included a number of patients who presented with relatively trivial diagnoses, as this reflects the referral patterns of our midwives, general practitioners and obstetricians within our hospital and local population. Our results show that all patients with specific dermatoses of pregnancy conformed well to the classification established by Holmes and Black in 1983. The role of the sex hormones [oestradiol, human chorionic gonadotrophin (hCG) and cortisol] in polymorphic eruption (PEP) and prurigo of pregnancy was studied in 125 cases and compared with 138 normal healthy pregnant controls. For pruritic folliculitis (PF), serum androgens were measured to establish if these were elevated. Nearly all patients were followed up postpartum, with respect to both maternal and fetal prognosis (some were unfortunately lost to follow-up). Many patients were primiparous (47%) and presented in their third trimester (49%). This study shows a surprisingly high prevalence of eczema during pregnancy. It is possible that earlier cases in the literature termed prurigo of pregnancy may in fact have been eczema, thus explaining the low incidence of prurigo in this study. Hormonal analysis showed a significant reduction in serum cortisol levels in patients with PEP compared with normal pregnant controls (P = 0.03), although hCG and oestradiol showed no differences. Serum androgens were not significantly elevated in patients with PF compared with controls. Birthweight (analysed by the individualized birthweight ratio) was significantly reduced in both the PF and pemphigoid gestationis groups. In the PEP and PF groups there was a male/female infant ratio of 2 : 1, not noted in previous studies. In all cases studied there were no adverse effects either on maternal or fetal outcome as a result of the pregnancy dermatosis. This study indicates that all patients fulfilled the criteria of the previous classification of the specific dermatoses of pregnancy, although we also now highlight the frequency of eczema in pregnancy and speculate as to possible causes. There were no cases of papular dermatitis of pregnancy. We feel that the specialist clinic is an important service which has improved the management of these women and identified areas for further research.