You’re on the edge of sleep. Then your leg kicks, or your whole body lurches, and you’re suddenly awake with your heart going — often with the distinct feeling that you were falling, or that you missed a step on a staircase that doesn’t exist. Nothing is wrong. Nothing woke you. Your own body did it.

Sleep researchers call this a hypnic jerk, or a sleep start. It’s benign and close to universal — a 2016 study published in Sleep Medicine puts the figure at 60% to 70% of the general population, and the true share is probably higher, since most of these jolts are too small to remember. They belong to a broad category of movements called myoclonus: brief, involuntary muscle contractions. Hiccups are myoclonus of the diaphragm . A hypnic jerk is the same basic event, fired once, at a very particular moment.

The Fall That Happens In The Handover Between Waking And Sleep

Falling asleep isn’t a switch being thrown. It’s a handover between two sets of circuits — arousal networks in the brainstem and hypothalamus that keep you alert, and sleep-promoting systems that shut them down. As that handover proceeds, the excitatory signal that keeps your muscles faintly tensed all day begins to withdraw, and muscle tone drops toward the near-paralysis of deep sleep .

That transition is not instantaneous, and for a few minutes the two systems overlap. The leading explanation for hypnic jerks is that in that window, a stray burst of activity in the motor pathways slips through before inhibition is fully in place, and reaches the muscles as a single hard contraction — a 2014 study published in Sleep Medicine was among the first to characterize the pattern of muscle activation in these jerks in detail, and a related 2016 study found no orderly spread of muscle activity from segment to segment — evidence, the researchers argued, that the signal originates deep in the brainstem rather than the cortex.

The falling sensation is the more interesting half. A widely held view among sleep researchers is that it isn’t the cause of the jerk but a story told afterward: a half-offline cortex receives a violent motor signal it can’t otherwise account for and constructs a plausible explanation — you were falling. A related idea points at the vestibular system, the balance apparatus in the inner ear. As muscle tone collapses, the brain may briefly misread the change as the beginning of a genuine fall and fire a righting reflex, the same automatic correction that saves you when you catch your toe on a curb.

Why Trees Might Help Explain This Fall

That misread-a-fall framing is what makes the arboreal hypothesis so appealing. If you sleep on a branch, losing muscle tone is not a metaphorical danger. A reflex that jolts the body, tests its grip and reasserts posture before sleep fully takes hold would be genuinely useful — and useful traits are what natural selection keeps. As Jason Ellis, a sleep scientist at Northumbria University, wrote in The Conversation , the jerk may let a sleeper test their footing before unconsciousness set in, particularly if that sleeper drifted off in a tree.

The ancestry is real enough. Primates spent tens of millions of years in the canopy, and most still sleep there. Monkeys sleep sitting upright on branches, cushioned by ischial callosities — tough, calloused pads on the rump. Great apes go further, building a fresh sleeping nest almost every night from bent and woven branches. Habitual sleeping on the ground is a recent development in our lineage, not the ancestral condition.

But the honest truth is that there is no direct evidence for the tree explanation, and it’s very hard to see how there could be. It’s a plausible narrative attached to a real phenomenon after the fact. Evolutionary biologists have a name for this temptation — the spandrels critique, laid out in a 1979 paper published in Proceedings of the Royal Society B by Stephen Jay Gould and Richard Lewontin, which argues that not every trait is an adaptation. Many are byproducts of how the underlying machinery is built. A hypnic jerk may simply be noise from an imperfect handover, no more designed than a hiccup.

There’s also a comparative problem. Dogs, cats and other mammals with no arboreal habit at all twitch at sleep onset. If the reflex were specifically about branches, you’d expect it to track tree-dwelling more closely than it does.

When This Fall Should Receive More Attention

Hypnic jerks get more frequent and more forceful under conditions that keep the arousal system running hot while the body pushes toward sleep — caffeine late in the day, high stress, accumulated sleep debt, intense exercise close to bedtime. That pattern fits the handover model well: the harder the two systems compete, the messier the transition.

They’re still harmless. The movements worth raising with a doctor are different ones — jerking that continues through established sleep or into the daytime, movements that cause injury, or confusion and disorientation afterward, which can point toward restless legs syndrome, periodic limb movement disorder or a seizure disorder rather than an ordinary sleep start.

What the jerk really shows is a seam. Sleep looks like one thing from the inside, but it’s assembled from separate systems that hand control back and forth — and every so often, on the way down, one of them lets go a fraction of a second before the other is ready.

Think a fake fall before sleep could really be a leftover survival instinct? Test how much you actually know about the forces that shaped us with this science-backed quiz: Evolution IQ Test