Three-month-old infants learned to activate a crib mobile by means of operant footkicks. Retention of the conditioned response was assessed in the presence of the nonmoving mobile. Although forgetting is typically complete after an 8-day retention interval, infants who received a reactivation treatment, a brief exposure to the reinforcer 24 hours prior to testing, showed no forgetting after retention intervals of either 2 or 4 weeks. Further, the forgetting function following reactivation did not differ from the original forgetting function. These experiments demonstrate (1) that "reactivation" or "reinstatement" is an effective mechanism by which early experiences can be maintained over lengthy intervals; and (2) that memory deficits in infants are best viewed as retrieval deficits. (Author)