Towards the end of a career that spanned almost 60 years, Ivan Petrovich Pavlov, eminent Russian physiologist, member of the Royal Society and Nobel prize-winner, concluded that sleep was a relatively straightforward process characterised by ‘spreading cortical inhibition’. By this Pavlov meant that as brain cells fatigue, they switch off one by one. When enough of them have switched off the brain falls asleep. 1 Conversely, as each cell becomes restored, then it switches back on and, when enough are switched on, the brain awakes. For Pavlov there was no single structure in the brain which controlled these events; the states of sleep and wakefulness were democratically selected by the cells of the cortex. While many scientists in the early decades of this century may not have agreed with the detail of this view, most shared the assumption that sleep was a fairly simple process during which individuals passed into a state of unresponsive somnolence, where they steadily remained until full consciousness returned. As all good scientists ought to be at some point in their career, Pavlov was wrong. It is now widely accepted that sleep is not only a complex, but also an active process. Far from ‘switching off, activity in the cortex (that part of the human brain responsible for so-called higher activities like thinking and solving problems) sometimes becomes quite intense even though the individual is sleeping soundly. Characteristics of sleep unrecognised by experienced scientists in the early 1920s have now become common scientific knowledge.