Maternal and infant health in primary care following childbirth is an important area which has recevied little research attention. Very little is known about primary health service use following childbirth and its determinants. The aim of this exploratory study was to test the hypothesis that place of birth is independently associated with mother and baby consultation patterns in the first year of life. This is a descriptive study of 453 pregnancies of 338 mothers from an inner city general practice between 1987 and 1996. There were 86 intended home births, 126 intended domino births and 240 intended hospital births, with one woman unbooked. Of these, 69 women had home births, 97 had domino births, 286 women had hospital deliveries and one baby was born in an ambulance. The outcome of interest was mother and baby consultation rates in the year following birth. In an attempt to control for pre-existing ill health we also calculated the women's consultation rate in the year prior to conception and substracted this from the mother and baby consultation rate in the year following birth. We analysed the data according to both intended and actual place of birth. Women intending a home birth had fewer mother and baby consultations in the year following birth compared to women intending domino births, who in turn had fewer consultations than women intending hospital care during labour (P < 0.001). This gradient persisted when pre-existing maternal ill health was accounted for. A similar gradient of increasing consultation rates through home births, domino birth, normal vaginal delivey in hospital and assisted delivery in hospital was demonstrated when data were analysed according to actual place of birth. Other maternal characteristics independently associated with increased mother and baby consultation patterns were: primiparity, manual occupational class, not breastfeeding, maternal smoking, infertility prior to pregnancy, and recorded anxiety, depression or life events before or after birth. Women intending and having a home birth were older (P < 0.001) and less likely to be primiparous (P = 0.055) or of manual occupational class (P < 0.001). Babies born at home were heavier (P < 0.001), and women experienced less perineal trauma (P < 0.001) and were more likely to breastfeed at birth (P < 0.001). The conclusion we draw from this cohort of births, is that there is an association between women intending a home birth and women who actually have home births and lower mother and baby consultation rates postnatally, when compared with women having normal vaginal domino or hospital births. This is independent of maternal age, parity, smoking during pregnancy, occupational class, a history of infertility or miscarriage in the year prior to conception, anxiety, depression or life events in the year prior to conception or after birth, breastfeeding, sex of baby and birth weight.