Animals do not all sleep for one long block at night, and they do not all pass through the same stages humans do. Depending on the species and setting, sleep may occur in many short bouts, while standing, underwater, or even during flight. Some birds and marine mammals can produce sleep-related slow waves in one brain hemisphere while the other remains more wake-like.
The bigger scientific challenge is deciding when these very different states should all be called sleep. Brain recordings work well for many mammals and birds, but researchers often rely on behavior, movement, heart rate, muscle tone, or neural activity in other animals. That means an animal sleep chart is not a simple ranking of equally measured facts.
How scientists recognize sleep in animals
In humans, sleep stages are identified with signals that include an electroencephalogram (EEG), eye movements, and muscle tone. Similar recordings reveal recognizable non-rapid eye movement (NREM) and rapid eye movement (REM) states in many other mammals and birds.
Those signals cannot be recorded easily in every animal, especially a tiny invertebrate, a fish swimming freely, or a large wild mammal. Comparative researchers therefore look for a cluster of behavioral features:
- a sustained period of reduced activity;
- a characteristic posture or resting place;
- a higher threshold for responding to a stimulus;
- rapid reversibility when the animal is disturbed; and
- a compensatory increase in sleep or sleep intensity after deprivation, called sleep rebound.
No single feature is decisive. An animal can be motionless while awake, and a sleeping animal can move. The combination of quiescence, reduced responsiveness, reversibility, and homeostatic regulation provides stronger evidence than posture alone 1.
Sleep is not the same as inactivity, torpor, or hibernation
Rest simply means activity has decreased. It does not establish that the brain is asleep. Torpor, hibernation, and estivation involve longer-lasting changes in metabolism and often body temperature that help an animal cope with cold, heat, drought, or limited food. Animals may sleep during parts of these states, but calling hibernation “a very deep sleep” erases important physiological differences 1.
Sedation, anesthesia, coma, and illness-related lethargy are also not normal sleep. This matters when interpreting an observation such as “the animal did not move for six hours.”
Why animal sleep patterns differ
Sleep takes time away from feeding, mating, parenting, migration, and watching for danger. The safest time and place to sleep also vary. A protected roost, burrow, social group, or position near the top of a food chain may give one species more opportunity to disengage from its surroundings than another.
Researchers have linked animal sleep with factors such as:
- time needed to find, eat, and digest food;
- exposure to predators while resting;
- whether a safe sleeping site is available;
- the need to breathe at the water's surface or remain in motion;
- body and brain development;
- temperature, daylight, season, and weather; and
- social, reproductive, and parenting demands.
These are tradeoffs, not universal rules. It is inaccurate to say that larger animals always sleep more, animals with faster metabolism always sleep less, or carnivores always sleep longer than herbivores. Comparative findings are associations across selected species, and different analyses have reached different conclusions about which ecological and biological traits matter most 2.
An individual animal's sleep can also change with age, food availability, migration, breeding, temperature, perceived danger, and whether it is in captivity. A species average can hide that flexibility.
Why animal sleep-duration charts can mislead
A number such as “12 hours per day” may come from a few captive animals under one light and feeding schedule. Another estimate may count immobility with a movement sensor, while another uses EEG and muscle recordings to separate quiet wakefulness, NREM sleep, and REM sleep. These methods are not interchangeable.
Laboratory and zoo settings reduce some demands of the wild, including foraging and predator vigilance. They can also introduce artificial light, noise, handling, unfamiliar social conditions, and restricted movement. A comparative review explains that sleep recorded in captivity may not represent sleep in an animal's natural ecological setting 3.
Two well-known field findings show why the method belongs beside the number:
-
EEG and muscle recordings from three adult female brown-throated three-toed sloths in a rainforest found an average of 9.63 hours of sleep per day, more than six hours below an earlier estimate from captive sloths. The studies differed in setting and in the ages of the animals, so the result does not establish one exact requirement for every sloth 4.
-
A study of two free-ranging African elephant matriarchs estimated about two hours of sleep per day from periods of trunk inactivity. The researchers explicitly described inactivity as a proxy that could not distinguish every period of quiet wakefulness from physiological sleep 5.
These studies are valuable because they bring measurement into the wild. They are not enough to crown a permanent “shortest sleeper” or rewrite a whole species from a handful of individuals.
Mammals: familiar stages with major adaptations
Most studied terrestrial mammals have NREM and REM sleep, but the duration, timing, depth, and arrangement of those states vary greatly. Many species distribute sleep across numerous bouts rather than one consolidated night 3.
Posture does not identify the stage by itself. Horses, for example, can have NREM sleep while standing or lying down. Their REM sleep occurs predominantly while recumbent, when the head and body can be supported during the associated loss of muscle tone 6. So “horses sleep standing up” is only part of the story.
Dolphins and other toothed cetaceans
In unihemispheric slow-wave sleep, one cerebral hemisphere shows more NREM-like slow-wave activity while the other remains more wake-like. The asymmetry later switches sides. This is not one half of the brain becoming completely inactive.
Toothed cetaceans studied with EEG, including dolphins and belugas, show this pattern. It is associated with asymmetric eye opening and can occur while the animal continues to move. Researchers think it helps reconcile sleep with surfacing to breathe, maintaining movement, caring for young, and monitoring the surroundings, although each proposed function is not equally established for every species 7.
Cetacean recordings have also found little or no unequivocal REM sleep in the form seen in terrestrial mammals. That does not mean cetaceans have no sleep. It shows that the mammalian sleep pattern familiar from humans is not mandatory for every mammal.
Birds can sleep with one hemisphere or both
Birds have NREM and REM sleep, generally arranged in shorter bouts than human sleep. Some can shift between bihemispheric sleep, asymmetric sleep, and unihemispheric slow-wave sleep. Keeping one eye open is associated with greater wake-like activity in the opposite hemisphere and can support visual monitoring 8.
Great frigatebirds provide a particularly clear field example. EEG recordings showed that the birds could use both unihemispheric and bihemispheric slow-wave sleep during flight. However, they spent far less time asleep and had shorter, lighter sleep bouts in flight than on land. They remained awake for most of the time even when gliding made sleep possible 8.
This does not show that birds can replace a full night's sleep with effortless midair naps. It shows that one species can temporarily reorganize and suppress sleep when ecological demands are high.
Reptiles do more than simply “rest”
Reptiles were once often described as having a simpler, uniform sleep state. Modern recordings have found more complexity. Studies in bearded dragons and Argentine tegus identified two recurring sleep states with some features that resemble mammalian and avian slow-wave and REM sleep.
The match is incomplete. Tegu sleep included brain oscillations not reported in the bearded dragon, and the REM-like state differed from mammalian REM in important ways. Researchers describe these as partial homologies, meaning that some traits may share an evolutionary origin while others have changed 9.
Eye movements or twitching in a sleeping reptile therefore do not prove that it is having human-like REM sleep or dreams.
Fish and invertebrates require different measurements
Fish do not need eyelids to sleep. Researchers identify sleep-like states through changes in movement, posture, responsiveness, and rebound, then add physiological or neural recordings when possible.
A 2019 study used brain-wide imaging alongside eye, muscle, and heart signals in larval zebrafish. It identified two neural sleep signatures, one with slow bursting activity and one initiated by propagating waves. The findings suggest deep evolutionary continuity in some vertebrate sleep mechanisms, but they do not mean a larval zebrafish passes through the same scored stages as an adult human 10.
Sleep-like behavior has also been documented in animals with very different nervous systems, including flies, worms, mollusks, and cnidarians. In these groups, scientists may not be able to use a scalp EEG or human stage labels at all. Shared behavioral criteria make comparison possible, while the differences prevent a one-to-one translation 1.
When animals sleep
Terms such as diurnal, nocturnal, crepuscular, and cathemeral describe when an animal is active. Diurnal animals are most active during the day, nocturnal animals at night, and crepuscular animals around dawn or dusk. Cathemeral species can distribute activity across day and night.
These are activity patterns, not separate kinds of sleep. They can shift with season, temperature, prey availability, predation, artificial light, human disturbance, and life stage. A nocturnal species is not necessarily asleep for every daylight hour, and a diurnal animal may nap during the day 2.
Do all animals sleep?
Sleep or sleep-like states have been found across remarkably distant branches of the animal tree. Even so, proving that every animal sleeps is difficult because the available tools and definitions change across species.
The bullfrog is often repeated as an animal that never sleeps. That claim came from an older study in which frogs remained responsive during periods of inactivity. It did not test the full modern set of sleep criteria, so it cannot establish that bullfrogs never sleep 11.
It is more accurate to say that no simple, verified list of “animals that do not sleep” exists. Some animals greatly reduce sleep during migration, mating, parenting, or food scarcity. Others show states that do not fit a human definition neatly. Limited detection is not proof of permanent sleeplessness 3.
What animal sleep can and cannot tell us about human sleep
Comparative research helps scientists identify conserved genes, brain circuits, and physiological processes. Animal experiments can also test questions that would be unsafe or impractical to study first in people.
But an animal finding is not automatically a human health result. A dolphin's unihemispheric sleep does not show that a person can train half the brain to sleep. A frigatebird's short sleep during flight does not prove that highly fragmented sleep is safe for people. A drug that changes sleep in a fly, fish, or mouse still needs suitable human studies before it can guide treatment.
Human sleep-duration advice and sleep-disorder care should come from human evidence. Animal behavior is evidence of what evolution has made possible in that species, not a sleep plan for another one.
FAQs
Which animal sleeps the least?
There is no secure all-species record. Estimates depend on how sleep was defined, whether the animals were wild or captive, how long they were observed, and how many individuals were studied. The two wild elephants in one field study averaged about two hours by an inactivity measure, but that does not establish the requirement of every elephant or exclude a lesser-studied species.
Which animal sleeps the longest?
There is no definitive winner here either. High estimates for bats, koalas, armadillos, and other animals often come from different methods and settings. A ranked list can create more certainty than the underlying studies support.
Can animals sleep standing up?
Yes. Some mammals and birds can enter at least some sleep states while standing. That does not mean posture makes them fully alert, or that they never need to lie down. Horses, for example, usually need a recumbent posture for REM sleep.
Do fish sleep if they cannot close their eyes?
Yes, many fish show well-supported sleep-like states. Eyelid closure is not required. Researchers look for reduced movement, characteristic posture, lower responsiveness, and sleep rebound, with neural or physiological measurements when available.
Do animals dream?
Dreaming is a subjective experience, so behavior cannot confirm what another species experiences. REM sleep, REM-like neural activity, twitching, or color changes may help researchers form hypotheses, but none directly reveals dream content.
Is hibernation the same as sleep?
No. Hibernation is an extended energy-saving strategy involving major metabolic and temperature changes. Sleep can occur within a hibernation cycle, but the two states are not interchangeable.
