Milk flow was recorded in 21 cows for three days after they were admitted to a large animal hospital. When the spontaneous flow of milk had stopped, a physiological dose (1 iu) of oxytocin was administered intravenously. Five of the cows were, in addition, treated with 0.35 mg of a long-acting analogue of oxytocin (carbetocin) one hour before the first milking after they were admitted. In the 16 cows not treated with carbetocin, only about 30 per cent of the total milk yield was released spontaneously on the first day, and the injection of 1 iu of oxytocin released approximately another 60 per cent of the total milk yield. On the second day, the proportion of the total milk yield released spontaneously increased and the fraction released after the injection of 1 iu oxytocin decreased. In contrast, the five cows treated with carbetocin released on average 94 per cent of the total milk yield spontaneously during the first milking.