Newborn infants are subject to repetitive painful and stressful events during neonatal intensive care. When the baby attempts to cope with a stressful situation the hypothalamus-pituitary-adrenal axis is activated, releasing cortisol. The free cortisol response is optimally measured in saliva and saliva samples can be taken easily and without pain. However, saliva is very scarce in infants and saliva stimulants can interfere with analytical methods. Nowadays, sweet solutions are frequently administered to neonates prior to a disturbing procedure in order to reduce pain. The possible interference of sweet solutions with the measurement of salivary cortisol has not yet been documented. The aims of the present study were to further improve the detection limit of the radioimmunoassay used for cortisol analysis and to determine the degree of interference of high concentrations of glucose with the analytical method. By decreasing incubation temperature and prolonging the incubation time it was possible to improve the detection limit of the radio immunoassay (RIA) to 0.5 nmol/L at the same time as the sample volume was decreased to 10 microL saliva. Saliva was collected from full-term and preterm babies and was sufficient for analysis in 113 out of 116 (97%) samples. Glucose in the concentrations and amounts commonly used for pain relief did not interfere with the RIA method. In conclusion, it is feasible to collect microlitre volumes of saliva and analyse even very low concentrations of cortisol in newborns. It is also possible to offer the baby oral glucose prior to a painful procedure and still reliably measure salivary cortisol.