Many sleep characteristics change acutely as early as during one’s thirties, while others change gradually throughout a person’s life span. We will discuss the cerebral mechanisms underlying age-related changes in sleep and their functional consequences. Identification of key mechanisms underlying age-related changes in Non-Rapid-Eye-Movement sleep (NREMS) oscillations: Our group put in evidence important changes in non-rapid-eye-movement (NREM) sleep oscillations during aging. We established that with increasing age, decreases in density (number of events per min of NREMS) and amplitude of slow waves (SW: >75uV and <4 Hz) and spindles (waxing and waning waves between 11 and 15Hz) occur predominantly in frontal regions, suggesting impairment in cortical circuits underlying NREMS oscillations. Indeed, our work suggested that these cortical neurons take longer to enter the SW hyperpolarization and depolarization phases in older than in younger individuals (Carrier et al., 2011; Martin et al., 2013). In addition, using Magnetic Resonance Imaging (MRI), we showed that cortical thinning in cerebral regions involved in SW generation explained age-related decreases in SW density and amplitude (Dube et al., 2015). Greater vulnerability with aging to challenged sleep-wake cycle: We will discuss how challenging conditions (such as shift work and jet lag) more easily disrupt sleep with aging. Hence, our group demonstrated that starting in the middle years of life (around age 40), older participants are less able to produce rebound SW during recovery sleep following sleep deprivation compared to younger individuals. In addition, we showed that older individuals have more difficulty maintaining their sleep when occurring at an abnormal circadian time, such as starting their sleep period in the morning. Our observations may explain why older subjects report more complaints associated with sleep deprivation, shift work and jet lag than younger individuals (Lafortune et al., 2012). NREM sleep oscillations and REM sleep are linked to cerebral and cognitive integrity in older subjects: Using simultaneous EEG and functional MRI (fMRI) recordings, we showed prominent age-related differences in functional connectivity (FC) across NREM sleep stages. These differences include lower FC decreases in cortico-cortical networks in the older individuals as compared to young individuals, as well as FC increases among anterior brain areas and between basal ganglia and various brain areas in the older group only. These modifications in cerebral FC may underlie age-related difficulty in sleep-induced brain plasticity. In addition, our work indicates that more REM sleep and higher spindle and SW density predict better performance on verbal learning, visual attention and verbal fluency (Lafortune et al., 2013). Moreover, in collaborative work, we used fMRI to determine the role of NREMS oscillations in motor learning in young and older participants. Our results indicated that the deficit in sleep-dependent motor memory consolidation in elderly individuals is related to a reduction in sleep spindle oscillations and to an associated decrease of activity in the cortico-striatal network. We also examined whether alterations in NREM sleep oscillations at a baseline visit were associated with increased likelihood of developing dementia at follow-up (4.5 years later) in patients with Parkinson disease (PD). Our results demonstrated that spindle alterations are associated with later development of dementia in PD and thus may serve as an additional predictor of cognitive decline in these patients (Latreille et al., 2015). Because REM sleep is regulated by the cholinergic system, which shows early degeneration in PD with cognitive impairment, we also evaluated whether REM sleep anomalies might mirror dementia development. Patients with PD who later developed dementia showed, at baseline, higher absolute power in delta and theta bands and a higher slowing ratio, especially in temporal, parietal, and occipital regions, compared to patients who remained dementia-free and controls (Latreille et al., 2016). Together, these results suggest that sleep EEG is a new promising predictive biomarker for PD dementia.