Mammalian hibernation is a natural process in which the brain undergoes profound adaptive changes that appear to protect the brain from extreme hypoxia and hypothermia. In addition to a virtual cessation of neural and metabolic activity, these changes include a decrease in adult neurogenesis; the retraction of neuronal dendritic trees; changes in dendritic spines and synaptic connections; fragmentation of the Golgi apparatus; and the phosphorylation of the microtubule-associated protein tau. Furthermore, alterations of microglial cells also occur in torpor. Importantly, all of these changes are rapidly and fully reversed when the animals arouse from torpor state, with no apparent brain damage occurring. Thus, hibernating animals are excellent natural models to study different aspects of brain plasticity. The axon initial segment (AIS) is critical for the initiation of action potentials in neurons and is an efficient site for the regulation of neural activity. This specialized structure-characterized by the expression of different types of ion channels and adhesion, scaffolding and cytoskeleton proteins-is subjected to morpho-functional plastic changes upon variations in neural activity or in pathological conditions. Here, we used immunocytochemistry and 3D confocal microscopy reconstruction techniques to measure the possible morphological differences in the AIS of neocortical (layers II-III and V) and hippocampal (CA1) neurons during the hibernation of the Syrian hamster. Our results indicate that the general integrity of the AIS is resistant to the ischemia/hypoxia conditions that are characteristic of the torpor phase of hibernation. In addition, the length of the AIS significantly increased in all the regions studied-by about 16-20% in torpor animals compared to controls, suggesting the existence of compensatory mechanisms in response to a decrease in neuronal activity during the torpor phase of hibernation. Furthermore, in double-labeling experiment, we found that the AIS in layer V of torpid animals was longer in neurons expressing phospho-tau than in those not labeled for phospho-tau. This suggests that AIS plastic changes were more marked in phospho-tau accumulating neurons. Overall, the results further emphasize that mammalian hibernation is a good physiological model to study AIS plasticity mechanisms in non-pathological conditions.