I'm a:
Client / Individual Therapist / Provider

The Science

REM sleep moves your eyes at 600 degrees a second. EMDR never asks for that.

Everyone says EMDR mimics the eye movements of dream sleep. We checked the speedometer. The claim doesn't survive the numbers — but what survives is more interesting.

Watch someone in deep sleep long enough and you'll catch it. Under closed eyelids, their eyes are darting around. Little twitches at first, then bursts — left, right, left, left, right. They're looking at something. Something that isn't there.

If you could strap a speedometer to those movements, here's what it would say: peak speeds of 400 to 600 degrees per second. Each individual jerk lasts 30 to 80 milliseconds. During the biggest ones, the eye sweeps across the entire visual field faster than you can say the word "look."

Now sit in on an EMDR session. The therapist's fingers travel left, right, left — roughly one full sweep per second. The client's eyes follow along, smooth and unhurried. That works out to maybe 20 degrees a second.

Not 400. Not close. Off by a factor of twenty, on a good day.

Two completely different kinds of eye movement

This isn't nitpicking over units. The darts of REM sleep are saccades — the ballistic jumps your eyes make when they snap from one point to another. The tracking in an EMDR session is smooth pursuit — the system you use to follow a bird across the sky.

They're run by different circuitry, and they can't impersonate each other. Pursuit physically can't go saccade-fast. Saccades can't go pursuit-slow; they're ballistic, which means once the brain fires one, its speed and landing spot are already baked in. Asking whether EMDR eye movements "match" REM eye movements is like asking whether a guided nature walk matches a sprint start. Both involve legs.

What REM eye movements are actually like

A few things worth knowing about the real thing, because they're stranger than the metaphor suggests:

So where did "EMDR mimics REM" come from?

The origin story is genuinely charming. In the late 1980s, psychologist Francine Shapiro went for a walk in a park, noticed that her own distressing thoughts seemed to lose their grip while her eyes moved back and forth, and built a therapy around the observation. Her new field needed an explanation for why waving fingers in front of someone could defuse a traumatic memory — and it reached for the nearest famous phenomenon involving eyes and processing: REM sleep.

As metaphors go, it's gorgeous. Sleep is when the brain files the day's emotional memories. EMDR appears to help the brain file memories that got stuck years ago. Same family, plausibly.

The problem is that somewhere along the way, "mimics REM" stopped being taught as a metaphor and started being repeated as a mechanism. And the moment you measure it, it falls apart. Nobody's eyes in an EMDR session are doing what eyes in dream sleep do. Not the speed, not the pattern, not the type of movement.

If it's not REM, what's doing the work?

The leading candidate is less poetic and better supported: working memory taxation. In a series of experiments largely out of the Netherlands — Marcel van den Hout, Iris Engelhard, and colleagues — people were asked to hold a disturbing image in mind while doing something attention-demanding at the same time. The result, replicated across many studies: the memory loses vividness and emotional punch. And here's the part the REM story never mentions — in lab conditions, counting tasks and tapping do the job too. Eye movements aren't magic. They're just a conveniently rhythmic way to keep working memory busy while a memory is held up to the light.

There's a second candidate: the orienting response — the built-in "what was that?" reflex that pulls attention to alternating sides and may keep the nervous system half-anchored in the safety of the present while the past replays. The honest position is that the mechanism is still debated, and anyone who tells you it's settled is overselling.

But give the REM intuition its due: it wasn't crazy, it was mislocated. Sleep researchers like Matthew Walker and Robert Stickgold have shown that REM sleep is deeply involved in emotional memory processing — in stripping the charge from the content. If EMDR plugs into the brain's memory-reconsolidation machinery, then "EMDR does something REM-like" may well be true at the level of what happens to the memory.

The REM story is a beautiful metaphor for what EMDR might do to a memory. It's a lousy description of what your eyeballs are up to.

Why the speed gap actually matters

This isn't just trivia for science night. It changes how you practice.

If you use self-guided BLS: you do not need REM speed. Don't whip your eyes back and forth as fast as they'll go — that's not what sleeping brains do anyway, and it's not what helps. Slow, comfortable tracking you can sustain for a full set is the target. The effect comes from holding the memory while your attention oscillates, not from velocity. If a session feels frantic, the pace is too fast.

If you're a therapist: set speed by the client in front of you, not by the metaphor. Some clients track best slowly; some tolerate faster sets. Watch the SUD drop across passes — that's your feedback loop, not a stopwatch.

Meanwhile, the real REM saccades remain one of the strangest things your body does: the fastest movements you'll make all day, performed nightly, in the dark, behind closed lids, seeing nothing. EMDR doesn't need to imitate that to be worth your attention. Whatever it turns out to be — memory, attention, taxation, some braid of all three — it's interesting enough on its own.

Feel the pace yourself

Try the free BLS tool

Visual, audio, and haptic bilateral stimulation in your browser — no login, no card. Set it to a slow, comfortable sweep and notice what a full set feels like.

Open the free BLS tool →