Sleep is traditionally defined and characterised by behavioural and electrophysiological criteria. For example, during sleep we are immobile and less responsive to the environment, and global cortical activity is distinctly different from an awake state. The differences between waking and sleep become less apparent as we look closer at the spatio-temporal patterns of cortical activity by recording local field potentials or neuronal spiking. It has been shown, that sleep-like patterns of neuronal activity are not uncommon during waking, even during active behaviours, and especially when the animals are drowsy or sleep-deprived. On the other hand, the main network oscillations during sleep – slow waves (~0.5-4 Hz) and spindles (~9-16 Hz) – are remarkably dynamic and idiosyncratic events, mostly occurring locally, and never encompassing the entire cortex at once. Slow waves are considered a reliable marker of preceding sleep-wake history, and a measure of sleep ‘intensity’. The ‘homeostatic principle’ postulates that the longer we stay awake, the more intense is our subsequent sleep. Recent evidence suggests that sleep homeostasis is a local process, and it has been identified both in cortical and subcortical structures, such as the dorsal striatum. Sleep spindles, which arise within the thalamocortical circuitry, also appear to occur locally in the neocortex; and their occurrence varies greatly depending on the cortical region, the time of day and the immediate preceding state. Finally, individual cortical neurons are highly diverse with respect to the state dependency of their spiking activity, and, importantly, their response to preceding sleep-wake history. Over the last few decades our knowledge about sleep has progressed tremendously. However, the fundamental questions remain: what is ‘noise’ and what is ‘signal’ in cortical activity during sleep, and how does the global and precisely regulated state of sleep emerge from the activity (or lack thereof) of local and distributed, cortical and subcortical circuits.