Paradoxically, fear conditioning (odor-0.5 mA shock) yields a learned odor preference in the neonate, presumably due to a unique learning and memory circuit that does not include apparent amygdala participation. Post-training opioid antagonism with naltrexone (NTX) blocks consolidation of this odor preference and instead yields memory of a learned odor aversion. Here we characterize the neural circuitry underlying this switch during memory consolidation. Experiment 1 assessed post-training opioid modulation of Fos protein expression within olfactory circuitry (olfactory bulb, piriform cortex, amygdala). Odor-shock conditioning with no post-training treatment (odor preference) induced significant changes in Fos protein expression in the granule cell layer of the olfactory bulb and anterior piriform cortex. Post-training opioid receptor antagonism (odor aversion) prevented the learning-induced changes in the anterior piriform cortex and also induced significant changes in Fos protein expression in the central nucleus of the amygdala. Experiment 2 assessed intra-amygdala opioid modulation of neonate memory consolidation. Post-training infusion of NTX within the amygdala permitted consolidation of an odor aversion, while vehicle-infused pups continued to demonstrate an odor preference. Overall, results demonstrate that opioids modulate memory consolidation in the neonate via modulating Fos protein expression in olfactory circuitry. Furthermore, these results suggest that opioids are instrumental in suppressing neonate fear behavior via modulating the amygdala.Odor-shock conditioning, a learning paradigm that readily produces fear in adult animals, produces a paradoxical odor preference in rat pups <10 d old (Camp and Rudy 1988;Sullivan et al. 2000a;Roth and Sullivan 2001, 2003Moriceau and Sullivan 2004a). This paradoxical behavior is attributable to a relatively simplistic olfactory learning and memory circuit that incorporates the olfactory bulb, anterior piriform cortex, and an apparent lack of amygdala participation (Sullivan et al. 2000a,b; Moriceau and Sullivan 2004a,b;Roth and Sullivan 2005;Moriceau et al. 2006). Indeed, we have demonstrated that the emergence of learned fear coincides with significant amygdala participation (Sullivan et al. 2000a;Moriceau et al. 2006). We suggest this attenuated aversion learning and potentiated preference learning early in development ensures pups readily learn to approach the maternal odor regardless of the quality of care she provides. Indeed, a recent study from our laboratory has documented this paradoxical odor preference learning within a seminatural nest environment as an odor paired with rough handling from a stressed mother resulted in a preference for that odor (Roth and Sullivan 2005). A similar paradoxical attachment despite caregiver abuse has been demonstrated in other species: chick (Hess 1962), infant dogs (Rajecki et al. 1978), and nonhuman primates (Harlow and Harlow 1965;Maestripieri et al. 1997Maestripieri et al. , 1999.We have previously demonstrated a ...