The mammary glands of rats either maintained continuously in the animal room on day 14 post partum or isolated in a room without other rats on day 21 post partum, refilled only partially within 8 h after the pups had withdrawn the milk from the glands by suckling with the aid of oxytocin injections to the mother. The partial refilling occurred regardless of whether or not the release of prolactin at the time of suckling was blocked by an injection of an extract of stalk median eminence (SME). The mammary glands of rats not injected with SME continued to refill and reached near capacity within 16 h whereas those of the SME-injected rats failed to refill any further. The glands of the SME-injected rats could be stimulated to refill more completely within 8 h after the initial emptying provided prolactin either was injected or was released endogenously by exposing the mothers to exteroceptive signals from their pups midway during the 8-h period of non-suckling which preceded the initial emptying. More complete mammary refilling also occurred within 8 h if a 4-h rather than an 8-h period of non-suckling preceded the initial emptying of the glands. The results of these investigations indicate that the prolactin released by suckling at the time of the initial emptying of the glands influenced only that milk which reaccumulated in the emptied glands during the 8th—16th h following suckling. The milk which reaccumulated during the first 8 h appeared to be influenced by prolactin released several hours before suckling. The results suggest, therefore, that a time delay of several hours exists from the time of release of prolactin by suckling until the full stimulatory effect of the hormone upon milk secretion is manifest. Thus the milk secreted after emptying of the glands at a given suckling would be the result of prolactin released earlier either at previous sucklings or in response to exteroceptive stimulation.