Do Dogs Dream? What Sleep Science Actually Shows

Published September 2026 · Pet N Pet Editorial · National Dog Day Science Series

Almost certainly, yes, though not in the way the internet usually tells it. Dogs cycle through the same broad sleep stages people do, including REM, the stage most linked to dreaming. In the foundational recording study, REM made up about 12% of a dog's sleep, with dream-stage episodes recurring roughly every 20 minutes (Lucas, Powell, and Murphree, Physiology & Behavior, 1977).

That much is measured. What a dog actually experiences during those minutes is a different question, and the honest answer is that no one has ever directly observed the content of a dog's dream. Here's what sleep science really supports, and where the popular claims run ahead of the evidence.

What Sleep Scientists Have Actually Measured in Dogs

Dogs move through the same architecture of sleep humans do: wakefulness, slow-wave (deep) sleep, and REM. In a 1977 study that recorded the brain, eye, and muscle activity of six pointer dogs across full 24-hour periods, sleep broke down to roughly 23% slow-wave sleep and 12% REM, with the rest spent alert or drowsy (Lucas, Powell, and Murphree, Physiology & Behavior, 1977). REM episodes averaged about six minutes and came around every twenty.

That study is the real source behind the everyday line that dogs "sleep like we do." It established that dogs have a genuine, recurring REM stage, the stage in which human dreaming is most vivid. It did not, and could not, tell us what a dog sees during that stage. It measured electricity and movement, not imagery.

More recent work adds the piece owners tend to care about: whether any of it matters. In a 2017 study, fifteen pet dogs learned a new command, then had their sleep recorded for three hours. They performed the task significantly better afterward than before (p=0.002), and the size of that improvement tracked their REM-sleep brain activity (Kis et al., Scientific Reports, 2017). So a dog's REM isn't idle noise. It's tied to locking in what the dog learned that day.

Why We Think Dogs Relive Their Day (and Why That's an Inference)

The popular claim that dogs dream about chasing squirrels or greeting their owner rests on a real and elegant body of research. The catch is that the research was done in rats, not dogs. In a 1994 study, neuroscientists recorded groups of "place cells" in the rat hippocampus while the animals ran a task. Cells that fired together during the task later fired together again during deep sleep, as if the brain were re-running the experience offline (Wilson and McNaughton, Science, 1994).

A follow-up went further. During REM specifically, rats replayed those waking activity patterns at close to real-time speed, stretches of behavior lasting tens of seconds to minutes, played back in the sleeping brain in the right order (Louie and Wilson, Neuron, 2001). The authors described it as a possible window into the content of dreams.

That's the strongest mechanism science has for the idea that a sleeping mammal relives its day. Dogs are mammals with the same basic sleep hardware, and the Kis findings show their REM is doing memory work too, so extending the idea to dogs is reasonable. But it is an extension. No one has recorded a dog's place cells replaying a walk in the park. When an article tells you a dog "is dreaming about you," that's a warm guess built on rat data, not a measurement.

What the Twitching, Paddling, and Soft Woofs Actually Are

The paddling legs, flicking whiskers, muffled barks, and eyes darting under closed lids nearly all happen during REM. In healthy sleep, the brainstem partly paralyzes the body during REM so the sleeper stays put, a mechanism called muscle atonia. The twitches you see are the small movements that leak past that paralysis.

How do we know the paralysis is what keeps a dog from acting out a dream? From a striking set of experiments in cats. When researchers disabled the brainstem region responsible for REM atonia, sleeping cats got up and performed coordinated behaviors, stalking, pouncing, grooming, while still asleep in REM (Jouvet and Delorme, 1965; Morrison et al., Journal of Sleep Research, 1993). It's the closest science has come to watching an animal enact a dream. Worth being precise, though: that work was in cats. Dogs share the same atonia system, which is why the finding is relevant, but the famous "pointer hunting in his sleep" image is an anecdote, not a lab result.

Do Small Dogs Really Dream More Than Big Dogs?

You'll see this stated everywhere as fact: small dogs supposedly dream every ten minutes or so, while big dogs dream less often but for longer. It's a tidy, quotable claim. It also has no primary source. It traces back to a popular-science writer's own extrapolation, and it isn't even present in his widely cited 2010 article on the subject. No polysomnography study has measured dream frequency by breed size.

What the sleep literature does support is an age pattern, not a size one. Puppies and older dogs tend to spend more of their sleep in REM than healthy adults, the same broad trend seen across many mammals. So if a puppy seems to twitch and paddle through half its naps, that fits the evidence. "A Chihuahua dreams six times as often as a Great Dane" does not. When you repeat a number, it's worth knowing whether anyone actually counted.

Should You Wake a Dog That Seems to Be Dreaming?

Usually, let it sleep. A dog deep in REM is disoriented for a moment on waking, the same groggy, half-here feeling you get if someone shakes you out of a vivid dream. A startled dog, especially an older or anxious one, can snap reflexively before it's fully awake, so "let sleeping dogs lie" is decent safety advice, not just a saying.

Twitching, soft woofs, and paddling are normal REM and not distress. If you genuinely need to rouse a dog that seems to be having a rough dream, use your voice from a small distance rather than your hands, and let it surface on its own. Signs that are worth a vet's attention are different from dreaming: rigid, prolonged full-body shaking, loss of bladder control, or an inability to wake, which can point to a seizure rather than a dream.

Frequently Asked Questions

Do dogs actually dream, or do they just twitch?

Both, and they're connected. Dogs have a genuine REM sleep stage, about 12% of their sleep in the foundational recording study, and the twitching, paddling, and soft woofs happen during REM, the same stage humans dream in. The physiology of dreaming is clearly there. What a dog experiences during it hasn't been directly observed.

What do dogs dream about?

The honest answer is that we don't know directly. The common idea that dogs replay their day comes from rat studies showing the brain re-runs waking experience during sleep. Since dogs share the same sleep machinery and use REM for memory too, it's a reasonable inference that they dream about familiar experiences, but it's an inference, not a measurement.

Why does my dog twitch, paddle its legs, or bark in its sleep?

Those movements happen during REM sleep. The brainstem normally paralyzes most muscles during REM so the sleeper stays still, and the twitches are the small movements that slip past that paralysis. Cat experiments that removed this paralysis caused sleeping animals to physically act out behaviors, which is strong evidence the sleeping brain is generating experience.

Do puppies and senior dogs dream more than adult dogs?

They likely spend more time in REM. Very young and very old animals across many species tend to have a higher proportion of REM sleep than healthy adults, so more visible twitching in a puppy fits the evidence. The separate claim that small breeds dream more often than large breeds, however, has no primary research behind it.

Should I wake my dog if it seems to be having a nightmare?

Usually not. A dog woken abruptly from REM is briefly disoriented and can snap reflexively, so it's safer to let it wake on its own. If you must, use your voice rather than touch. Prolonged rigid shaking, loss of bladder control, or an inability to wake are different from dreaming and warrant a vet check.

Bottom Line

Dogs dream in every sense science can actually test. They have real REM sleep, their sleeping brains do measurable memory work, and the shared mammalian atonia system explains the twitching and the fact that they normally stay put while doing it. The part we can't yet see is the picture in their heads. The lovely idea that your dog is dreaming about you is a fair bet, built on rat brains and a bit of faith, not a settled fact. Either way, the paddling paws on the couch are the real thing.

SOURCES & CITATIONS

Lucas, Powell, Murphree "Baseline sleep-wake patterns in the pointer dog," Physiology & Behavior 19(2):285-291 (1977). Foundational canine polysomnography (6 dogs): ~23% slow-wave sleep, ~12% REM. https://pubmed.ncbi.nlm.nih.gov/203958/

Kis, Szakadát, Gácsi, et al. "The interrelated effect of sleep and learning in dogs (Canis familiaris)," Scientific Reports 7:41873 (2017). 15 dogs; post-sleep task gain p=0.002, tracked REM EEG. https://pmc.ncbi.nlm.nih.gov/articles/PMC5292958/

Wilson, McNaughton "Reactivation of hippocampal ensemble memories during sleep," Science 265(5172):676-679 (1994). Rat hippocampal replay during sleep. https://pubmed.ncbi.nlm.nih.gov/8036517/

Louie, Wilson "Temporally structured replay of awake hippocampal ensemble activity during REM sleep," Neuron 29(1):145-156 (2001). Rat replay at near-real-time speed in REM. https://pubmed.ncbi.nlm.nih.gov/11182087/

Morrison et al. "Mechanisms underlying oneiric behaviour released in REM sleep by pontine lesions in cats," Journal of Sleep Research 2(1) (1993); with Jouvet & Delorme (1965). Dream-enactment work in cats. https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1365-2869.1993.tb00052.x

Iotchev, Kis, Bódizs, et al. "EEG Transients in the Sigma Range During non-REM Sleep Predict Learning in Dogs," Scientific Reports 7:12936 (2017). Dog sleep spindles linked to memory (citation-level verification). https://www.nature.com/articles/s41598-017-13278-3

```