In a retrospective questionnaire survey of 885 women who had given birth 3-5 months before, fewer of those who were still breast-feeding at 4 months (n = 645) were using drugs than those who had stopped breast-feeding before 4 months (n = 240), during the 2 week period preceding registration. The average number of doses (Defined Daily Doses/1000 women/day) was 166 and 307, respectively, in that period. The number of doses taken was significantly associated with the use of oral contraceptive agents (p less than 0.005) and young maternal age (p less than 0.05). Most of the variation in drug use between breast-feeding and not breast-feeding mothers was probably due to the greater use of contraception by the latter. The number of drugs used per mother in the 4 month period seemed to be best predicted by her and her infant's disorders. Long-term medication in breast-feeding women included many drugs for which there is incomplete or no data about milk transfer, e.g. salbutamol, clemastine, dexchlorpheniramine, phenylpropanolamine, cromoglycate and levomepromazine. The disorders most extensively treated with drugs in this period were dyspepsia, haemorrhoids and inflammation of the breast. The finding that smoking was associated with early weaning and consumption of alcohol with prolonged breast-feeding calls for further investigation. More information on these drug and health issues to the breast-feeding mother is highly desirable.