The mutants showed impaired vestibulo-ocular reflex as well as impaired acquisition of classical delay conditioning of their eye blink response. In the present work, we found that learning-induced LTP was absent in subjects that received muscimol into BLA before training. Likely, this effect is due to an interference with LTP induction occurring during CS-US presentation. However, BLA is necessary also for pain-related response and for the regulation of fear innate behavior. Thus, although we did not observe a significant change in animals’ spontaneous activity before fear acquisition, we cannot exclude an effect on CS and/or US processing produced by pretraining BLA inactivation. In line with the present findings, previous studies reported that pretraining BLA blockade attenuated activity-dependent processes in thalamus, cingulated cortex, and hippocampus. In a second line of experiments, we blocked BLA after learning, i.e. during the consolidation phase of memory process. This approach allows us to rule out any interference with sensory or painful stimuli processing so that any effect on cerebellar plasticity is only due to the interference with the memory trace. To date, the only study that tested the effect of BLA inactivation on activitydependent processes that occur in regions engaged in consolidating long-term memories has been LY2109761 performed by McIntyre et al.. The authors showed that post-training infusion of lidocaine into BLA significantly reduced the increase in Arc protein observed in hippocampus following avoidance learning. Our study extends these results to the long-term synaptic plasticity, i.e. LTP, which underlies memory formation in the cerebellar cortex. In the hippocampus and cerebellum the electrically-induced LTP is widely considered a cellular model of learning. A support to this hypothesis comes from recent findings showing that learninginduced LTP interferes with the subsequent electrically-induced LTP in hippocampus and cerebellum. The present results, however, reveal an important difference between electrically- and learning-induced LTP. Namely, the latter type of LTP requires information from other regions to be formed and maintained. Hebbian model of learning maintains that pre- and postsynaptic neurons have to be coactive within a defined time period to modify synaptic strength. In our model, CS and US reaching the cerebellar cortex produce LTP provided that a heterosynaptic input coming from BLA sets the proper local conditions of such an interaction. Thus, studies employing the electrically-induced LTP in order to identify the cellular mechanisms related to memory processes should take into account the heterosynaptic inputs, considering them as integrative units. Theoretically, the functional meaning of the heterosynapticdependence of learning-induced LTP might be that local synaptic processes underlie the automatic recording of an attended experience.