Caffeine, which is used for the treatment of apnoea in premature newborns, is known to be excreted into breast milk. However data on the amount of caffeine transferred to the breast-fed infant and on caffeine concentrations in the baby are lacking. In 18 healthy breast-feeding women caffeine concentrations in breast milk were measured 2 and 4 hours after the intake of coffee (145.8 mg caffeine, mean +/- sd, n = 18). For an estimation of kinetic parameters (eg, AUC), additional saliva samples were collected up to 6 hours after coffee intake. The daily caffeine intake of the infants was calculated from the average breast-milk concentration (AUCsaliva X milk/saliva ratio/24 hours) as average milk concentration X daily milk volume. From nine of the babies (aged 20 days to 19 weeks) at least one saliva sample could be obtained. The ratio milk/saliva was found to be 0.90 +/- 0.20 (mean +/- sd, n = 18) and the average breast-milk concentration was 0.82 +/- 0.29 mg/L (mean +/- sd, n = 18). The daily caffeine intake of the infants was calculated to range from 0.027 to 0.203 mg/kg/day. The caffeine concentrations measured in the babies ranged from less than 0.05 to 0.75 mg/L. Hence it can be concluded that the amount of caffeine ingested by the children is small compared to the therapeutic dose if usual amounts of coffee are taken by the mothers.