A heterodyne pulsed doppler velocitymeter was used to measure blood velocities in the mammary branch of the lateral thoracic artery during breast-feeding and after oxytocin injections. A few heart beats before the mother felt the milk-ejection reflex, blood velocities fell rapidly by 40-50%, and then increased during the next 1-2 min. Synthetic oxytocin was given intravenously to lactating women, both in doses believed to be physiological in milk-ejection and larger, and the same velocity changes occurred. Synthetic oxytocin was also given to non-lactating women who had never been pregnant or who had breast-fed previously and to one man. In these a rapid increase in velocity was observed after about a half a minute, lasting for 1-2 min. In addition to the short-term effects of breast-feeding, blood velocity increased rapidly towards the end of nursing or a few minutes afterwards, with a maximal increase of 25-50%. It then slowly returned to pre-nursing values during the following 30-60 min. This pattern was seen in both breasts, independent of which breast the infant had been sucking.