While 3-month-old infant rhesus macaques (Macaca mulatto) were awake and active in social interactions away from their mothers, body and nipple contacts with their mothers were nevertheless made from time to time. In each dyad the proportions of contacts made by the mother nearly equalled those broken by her, suggesting a meshed interaction in which each partner accepted most of the other’s contact initiations and terminations. Passive prevention of nipple contact by a mother reduced the frequency of nipple contact by her infant in the first 5 s after the infant had made body contact. Passive prevention occurred after fewer than 1 in 6 body contacts initiated by infants, and – even without its occurrence – most infants were less ready to take the nipple after their own initiatives than after maternal initiatives. Once nipple contact had been made, the probability of breaking body contact was reduced. The role of maternal rejection both in the control of nipple contact in the short term and in determining (through its effect on the sucking pattern) whether the mother gives birth in the next birth season or later is discussed. We suggest that, by the age of 3 months, the infants had already learned when and how often nipple contact with their mothers would be acceptable during their awake and active periods, and we suggest that subsequent decreases in the frequency of nipple contact were partly the results of maternal rejections which were accepted by the in