The mammary pheromone promotes the acquisition of novel odorants (CS1) in newborn rabbits. Here, experiments pinpoint that CS1 becomes able to support neonatal learning of other odorants (CS2). We therefore evaluated whether these first-and second-order memories remained dependent after reactivation. Amnesia induced after CS2 recall selectively blocked this memory, when recall and amnesia of CS1 left the souvenir of CS2 safe; this finding partially differed from results obtained in adult mammals. Thus, in this model of neonatal appetitive odor learning, second-order memory seems to depend on first-order memory for its formation but not for its maintenance.The mammary pheromone (MP) emitted by lactating rabbit females helps newborn rabbits to orient to the mother, then rapidly to localize and grasp the nipples (Coureaud 2001;Schaal et al. 2003). It is a determinant for survival and growth in the context of mother -young interactions daily limited (,5 min) (Zarrow et al. 1965). The MP also acts as a cognitive organizer promoting the rapid (single trial) appetitive learning of new odor stimuli, initially neutral. After pairing with the MP, the learned stimulus gains significance (i.e., releases the sucking behavior of neonates) in less than 24 h, following rules of first-order conditioning where the MP constitutes the unconditioned stimulus (US) and the novel odorant the conditioned one (CS1) (Coureaud et al. , 2008(Coureaud et al. , 2009.Consolidation and reconsolidation processes, usually required in long-term memories (McGaugh 1966(McGaugh , 2000Nader et al. 2000b;Sara 2000), are also involved in MP-induced odor memory; retention is impaired by injection of a protein synthesis inhibitor (anisomycin), and anisomycin induces amnesia also after reactivation of the previously consolidated memory (Coureaud et al. 2009). Nevertheless, properties of learning and memory promoted by the MP remain poorly known. The present experiments aimed to go further, to determine whether secondorder odor learning exists in this model (CS2-CS1 pairing) and to examine whether first-and second-order memories remain dependent in the neonatal brain. To date, Debiec et al. (2006) showed in adult rats that following first-, then second-order conditioning, CS2 memory remains dependent of CS1 memory for its expression in the long-term. Since mechanisms which support learning present specificities in newborns (Wilson and Sullivan 1994; Languille 2010), we suspected that early memories could be less dependent on each other after formation based on chainlike associations.We investigated these early memory processes (n ¼ 115 neonates, from 32 females) and their sensitivity to the action of anisomycin (AN; 42 mg/kg i.p.; Aldrich) after reactivation. Like other authors, using other species-e.g., adults (for review, see Davis and Squire 1984;Desgranges et al. 2008;Merhav and Rosenblum 2008) and newborns (Gruest et al. 2004;Languille et al. 2008Languille et al. , 2009, we defend that the blocking effect of AN on odor memory in newborn rabbits cannot ...