Translational research on Alzheimer's disease (AD) has often focused on reducing the high cerebral levels of amyloid-β (Aβ) as a key characteristic of AD pathogenesis. There is, however, a growing body of evidence that synaptic dysfunction may be crucial for the development of the most common (sporadic) form of AD. The applicability of melatonin (mainly produced by the pineal gland) to the treatment of AD is actively evaluated, but usually, such studies are based on animal models of early-onset AD, which is responsible for only ~5% of AD cases. We have shown previously that in OXYS rats (an established model of sporadic AD), accumulation of toxic forms of Aβ in the brain occurs later than does the development of signs of neurodegenerative changes and synaptic failure. In this regard, recently, we uncovered beneficial neuroprotective effects of melatonin (prophylactic dietary supplementation) in OXYS rats. Our aim here was to evaluate, starting at the age of active progression of AD-like pathology in OXYS rats, the effects of long-term oral administration of melatonin on the structure of synapses and on neuronal and glial cells of the hippocampus. Melatonin significantly increased hippocampal synaptic density and the number of excitatory synapses, decreased the number of inhibitory synapses, and upregulated pre- and postsynaptic proteins (synapsin I and PSD-95, respectively). Furthermore, melatonin improved the ultrastructure of neuronal and glial cells and reduced glial density. Based on our past and present results, the repair of neuroplasticity by melatonin is a promising strategy against AD.