The only social cat. A pride is a group of related females — the males arrive as a coalition and hold it for a few years — and lions are not social because hunting together works, which was tested and does not.
Lions are the only cats that live in permanent groups, and the reason is not the one everybody gives. Cooperative hunting is the obvious explanation and it was tested directly in the Serengeti over years: measured as food per lion, a lioness in a group frequently does worse than one hunting alone. What group living actually buys is the ability to hold territory against neighbouring prides and to defend cubs against males, and larger female groups win encounters, take the better ground — river confluences where prey concentrate — and raise more surviving cubs. The competition that shapes lion society is with other lions. That reframes the pride itself. The stable core is a group of related females who stay in their natal range for life; the males are the transient element, dispersing as adolescents, forming coalitions with brothers and cousins or sometimes with unrelated males, and holding a pride for perhaps two to three years before being displaced. When a new coalition takes over it kills the unweaned cubs, which returns females to breeding condition sooner — and the synchronised births, communal nursing and collective cub defence that characterise prides are counter-adaptations to exactly that. Several other familiar claims need the same treatment. Males hunt regularly and successfully, in cover and at night, where the classic open-plains daylight studies could not see them. Lions scavenge heavily, including from hyenas, and carcasses move in both directions depending on numbers. The mane advertises condition rather than protecting the neck, and its cost is heat — which is why males in the hottest regions have short pale manes or none. And the twenty-hour figure everyone quotes records inactivity rather than sleep, because nobody has measured sleep in a wild lion.
Developed record · 90% complete · reviewed 2026-08-11
What this page covers
One species across sub-Saharan Africa plus a single small population in Gir, India. Modern treatments recognise two subspecies: Panthera leo leo, covering West, Central and North African populations together with the Asiatic lions, and Panthera leo melanochaita, covering East and Southern Africa.
Often confused with: Mountain lions, which are cougars — a different genus entirely and not closely related; Ligers and tigons, which are captive hybrids with no wild populations
Quick facts
Why they live in groups
Territory and cub defence — not hunting, which was tested
The pride
Related females who stay; male coalitions hold it for 2–3 years
Do males hunt
Yes — in cover, at night, favouring larger prey
The mane
A condition signal; its cost is heat, not fighting
It was tested in the Serengeti over years. Lions in groups eat less.
Lions are not social because hunting together works better — it measurably does not
Established
Specialists would state this without hedging. Multiple independent lines of evidence agree.
Per-capita food intake in Serengeti lions does not increase with group size across most of the observed range, and lionesses in medium-sized groups frequently obtain less food than singletons or pairs. Female group size instead tracks the ability to defend a territory against neighbouring prides and to protect cubs from infanticidal males. Larger female groups win intergroup encounters, acquire and retain higher-quality territories — particularly river confluences — and produce more surviving offspring.
Who this applies to
Serengeti lions, from decades of continuous observation
Studied in
Panthera leo
You may have heard
“Lions live in prides because hunting together lets them take bigger prey”
It is the obvious explanation, it was tested, and it failed. Measured over years, a lioness in a group frequently eats *less* than one hunting alone. What group living actually buys is the ability to hold territory against other prides and to defend cubs against incoming males. The competition that shapes lion society is with other lions, not with prey — which is why territory maps predict reproductive success better than hunting success does, and why lions hunt cooperatively rather badly.
Why we rate it this way, and what the caveats are
EstablishedHigh confidence
The obvious explanation was tested directly against measured intake over years and failed, and the alternative was subsequently supported by territory mapping linked to reproductive success across decades of individual records.
How far it can be extended
The territorial and infanticide pressures are general to lions, though prey abundance in the Serengeti is unusually high and may weaken feeding benefits relative to other systems.
Caveats
One ecosystem, albeit exceptionally well characterised over decades.
Per-capita intake is difficult to measure precisely and rests on observed kills and estimated shares.
Correlational for group size, which cannot be manipulated experimentally.
Still unanswered
Does the same conclusion hold in systems with lower prey density, where feeding benefits of cooperation might be larger?
Larger groups winning encounters, holding better territory, and producing more surviving offspring — the mechanism that does explain it.
How we know
Testing the explanation everybody already believed
Do lions live in groups because hunting together works better?
Lions are the only social cat, and the explanation was obvious enough that it went largely unexamined for decades: hunting together lets them take larger prey. It makes a prediction that can be checked, though, and the check is unglamorous. If cooperative hunting explains sociality, then a lion in a group must end up with more food than a lion alone. So food intake per lion was measured — not hunting success, which is a different quantity, but how much each animal actually ate — across Serengeti prides of different sizes, over years, using individually known lions in a population followed continuously since the 1960s.
What happened
Per-capita intake does not rise with group size across most of the observed range, and lionesses in medium-sized groups frequently do worse than singletons or pairs. Group size instead tracks the ability to hold a territory and to defend cubs against incoming males.
What it shows
That the standard explanation for lion sociality is wrong, and it shows it by measuring the quantity the explanation predicts rather than the one that is easier to observe. It also makes sense of something otherwise puzzling: lions hunt cooperatively rather badly, with individuals frequently holding back — which is unsurprising if the group is not there for hunting.
What it does not show
One ecosystem, and an unusually prey-rich one, so feeding benefits of cooperation may be larger where prey is scarcer. Per-capita intake is hard to measure precisely in the field and rests on observed kills and estimated shares. And group size cannot be manipulated, so the alternative explanation is supported by correlation across decades rather than by experiment.
The controls — what makes this evidence rather than a story
Per-capita intake measured rather than hunting success rate, since a group that catches more may still feed each member less.
Individually identified lions across many prides, so group size is a real variable rather than an average.
Years of observation, covering seasonal and annual variation in prey.
The full range of group sizes, including singletons and pairs, rather than comparing only large groups against each other.
Lions are the only social cat, and for decades the explanation went unexamined because it seemed self-evident: hunting together lets them take larger prey. It makes a testable prediction, though — a lion in a group must end up with more food than a lion alone — and the prediction fails. Measured as intake per lion across Serengeti prides over years, group living does not pay. Lionesses in medium-sized groups frequently do worse than singletons or pairs.
It also explains something otherwise puzzling. Lions hunt cooperatively rather badly: individuals hold back, let others commit first, and take a share regardless. That is difficult to reconcile with a group organised around cooperative hunting, and unsurprising in a group organised around something else.
What group living does buy is territory and cubs. Larger female groups win encounters with smaller ones, acquire and hold better ground — particularly river confluences, where prey concentrate — and raise more surviving offspring. Territory quality rather than group size alone predicts reproductive success. The pressure shaping lion society is other lions.
Neighbouring prides are the main threat a pride faces, and intergroup encounters are a significant cause of death. A lion is more likely to be killed by another lion than by anything else.
How much food each individual actually gets, as opposed to how much the group catches. The distinction is what overturned the cooperative-hunting explanation.
A pride is a group of related females; the males are visitors who hold it for a while
Established
Specialists would state this without hedging. Multiple independent lines of evidence agree.
A lion pride comprises related adult females and their dependent offspring, which forms the stable core, together with a coalition of one or more adult males whose tenure is temporary. Males disperse from their natal pride as adolescents and form coalitions, most often with brothers and cousins but sometimes with unrelated males; larger coalitions take over prides more successfully and hold them longer. In small coalitions reproduction is shared relatively evenly; in large ones it is more skewed. Incoming males kill unweaned cubs, which returns females to reproductive condition sooner, and females show counter-adaptations including birth synchrony, communal nursing in crèches, and mating during periods when conception is unlikely.
Who this applies to
lions generally, established chiefly from long-term Serengeti records
Studied in
Panthera leo
You may have heard
“A pride is ruled by one dominant male — the king — with his females”
The stable part of a pride is the females, who are related to each other and stay for life; the males are the transient element, arriving as a coalition and holding the pride for perhaps two to three years before being displaced. So the ownership runs the other way round from the story. And the takeover is not a change of management: incoming males kill the cubs, which is why female lions synchronise births, nurse each other's cubs and defend them collectively. Leaving that out makes everything else about pride life incomprehensible.
Why we rate it this way, and what the caveats are
EstablishedHigh confidence
Individual lions have been followed continuously in the Serengeti since the 1960s, so coalition formation, takeover, tenure and cub fate are recorded rather than inferred.
How far it can be extended
Female philopatry, male dispersal into coalitions, takeover and infanticide are documented across African lion populations, with Asiatic lions differing in that males associate with females only around mating.
Caveats
Most detail comes from one intensively studied population.
Relatedness in the early work was inferred from observed maternal history rather than genotyping.
Asiatic lions differ in social structure and are a small, separately managed population.
Still unanswered
What determines whether an unrelated male is accepted into a coalition, given that reproduction is then shared with a non-relative?
The founding description of pride structure, territoriality and activity budgets that later individual-level work refined.
The stable part of a pride is the females. They are related — mothers, daughters, sisters, aunts — and they stay in their natal range for life. Everything durable about a pride, including its territory and its identity, belongs to them.
The males are the transient element, and this is the piece the popular picture inverts. Young males disperse as adolescents, spend a nomadic period, and form coalitions — usually with brothers and cousins, sometimes with unrelated males they meet while nomadic. A coalition then takes over a pride and holds it for perhaps two to three years before a larger or younger coalition displaces it.
Two very different lives
Aspect
Females
Males
Where they live
Natal range, for life
Disperse as adolescents, never return
Group
Related kin group, stable across generations
Coalition, formed in the nomadic period
Tenure in a pride
Permanent
Typically two to three years
Main threat
Neighbouring prides
Larger coalitions, and the nomadic period itself
On losing
Nothing to lose — they stay
Displaced, rarely take another pride
Coalition size is the currency. Larger coalitions take prides more successfully and hold them longer, which is why males form them at all — and why unrelated males cooperate despite sharing no genes. A coalition of two beats a coalition of one whether or not the partners are brothers, so the partnership pays for itself. In small coalitions reproduction is shared fairly evenly; in large ones it becomes more skewed.
Which makes "the king of the pride" a poor description in two directions at once. There is frequently more than one male, they are usually there for a couple of years, and the group they have joined was there before them and will be there after.
Words used here
Coalition
A group of adult male lions that disperse, roam and take over prides together. Usually relatives, sometimes not.
Philopatry
Staying in the area where you were born. Female lions are philopatric; males are not, which is why prides are female kin groups.
Takeover
The displacement of a resident coalition by another. Happens every few years and is the central disruptive event in pride life.
The part documentaries skirt, and without which the rest makes no sense.
When a new coalition takes a pride, it kills the unweaned cubs. A female nursing another male's cubs will not conceive for many months, and a coalition holds a pride for only two or three years — so from the incoming males' position, waiting is the same as not breeding at all.
Recording this is not gratuitous. It is the pressure the rest of pride life is organised around, and a benign account that leaves it out has removed the explanation for everything else. Female lions synchronise their births so that cubs across the pride are of similar age and can be defended together. They form crèches and nurse each other's cubs. They mate with incoming males during periods when conception is unlikely, which appears to confuse paternity and buy time. Related females defend cubs collectively and sometimes successfully.
None of those behaviours reads as general sociability once the threat is visible. They are answers to a specific recurring event, and they are why a pride of related females is a more effective arrangement than a group of unrelated ones would be.
Cubs that survive it face a long dependency. They nurse for six to eight months, remain dependent for around two years, and mortality before adulthood is high — often more than half. Females stay; males leave at two to three years old and enter the nomadic period, which is the most dangerous stretch of a male lion's life.
Animal welfare
On watching this happen
Infanticide following a takeover is distressing to watch and is a normal part of lion biology rather than an aberration or a sign of a disturbed population. Wildlife authorities and reserve managers do not intervene, and the reasons are sound: the behaviour is the mechanism by which lion populations are structured, and intervening would alter the population in ways nobody can predict. If you encounter it on a guided drive, the guide has seen it before.
Where this applies: global
Words used here
Crèche
A group of similarly aged cubs raised together by several related females, who nurse each other's young. A direct consequence of birth synchrony.
Infanticide
Killing of unweaned young by incoming males. Documented across lions and several other mammals with the same social structure.
Male lions hunt regularly — in cover, at night, where researchers historically could not watch
Well supported
Good evidence backs this, though some details remain open.
Systematic observation of hunts in the woodland habitat of Kruger National Park found male lions hunting substantially more than the standard account allows, and hunting successfully. Male hunts are concentrated in dense cover and at night, where stalking to close range is possible, and target large prey such as buffalo more frequently than female hunts do. Female hunts are more often in open habitat and in groups. The prevailing view that males do not hunt derives from earlier work conducted in open plains during daylight, where a large conspicuous animal cannot stalk effectively.
Who this applies to
measured in Kruger woodland, against earlier open-plains observation
Studied in
Panthera leo
You may have heard
“Female lions do all the hunting while the males eat what they catch”
This is a sampling artefact that became a fact about lions. The classic studies were done on open plains in daylight, because that is where researchers could see — and in the open, in daylight, a male with a mane genuinely cannot stalk. Watch in woodland at night and males hunt regularly and successfully, favouring larger prey than females take. "Males do not hunt" described the observation conditions rather than the animals.
Why we rate it this way, and what the caveats are
Well supportedHigh confidence
Direct observation in a habitat where earlier work had not looked, producing a result that explains the earlier finding rather than merely contradicting it.
How far it can be extended
The habitat explanation predicts that male hunting frequency should track cover availability, which is consistent with observations across sites of differing habitat.
Caveats
Night and dense-cover observation remains harder, so male hunting may still be under-recorded.
Kruger woodland differs from Serengeti plains in ways beyond visibility.
Separating hunting from scavenging and kill appropriation is difficult in some records.
Still unanswered
How much of the variation in male hunting between populations is habitat and how much is coalition size?
Establishes the preferred prey size band across sites, within which the male preference for larger prey sits.
"Females do all the hunting" is a sampling artefact that became a fact. The classic lion studies were done on open plains in daylight, because that is where researchers could see — and in the open, in daylight, a male with a mane genuinely cannot stalk. Observation in woodland at night found males hunting regularly and successfully, favouring larger prey such as buffalo. The claim described the observation conditions rather than the lions.
They specialise on large prey — and steal a great deal of it from hyenas
Established
Specialists would state this without hedging. Multiple independent lines of evidence agree.
Compiled kill records across African study sites show lions preferentially taking prey in a band of approximately 190 to 550 kilograms, with a most-preferred mass near 350 kilograms, and avoiding much smaller and much larger species relative to availability. The preference holds across sites with differing prey communities. Lions also scavenge substantially, including taking carcasses from spotted hyenas; carcass transfer occurs in both directions and depends on the relative numbers of each species present.
Who this applies to
African lions across compiled study sites
Studied in
Panthera leo, Crocuta crocuta
You may have heard
“Lions are noble hunters and hyenas are scavengers that steal their kills”
This has been backwards since it was first checked in the 1960s, and the correction is still not widely known. Lions scavenge a great deal, including taking carcasses from hyenas — which hunt successfully and often. Transfer runs both ways and is settled largely by numbers: enough hyenas will displace a lion, and a coalition of males will displace any number of hyenas. The moral framing came from documentaries rather than from field records.
Why we rate it this way, and what the caveats are
EstablishedHigh confidence
A cross-site compilation in which the same preference emerges from independent prey communities, supported by decades of direct observation of carcass acquisition.
How far it can be extended
Prey preference was consistent across sites with substantially different prey communities, which is what distinguishes a predator property from local availability.
Caveats
Kill records under-represent small prey, which are consumed quickly and rarely observed.
Compiled from studies with differing methods and detection biases.
The direction of carcass transfer between lions and hyenas varies with local densities of each.
Still unanswered
What proportion of a lion's intake is scavenged rather than hunted, across systems with different scavenging opportunity?
The founding documentation of lion scavenging, including from hyenas, and of carcass transfer in both directions.
What lions take is unusually consistent. Compiling kill records across African sites gives a preferred prey band of roughly 190 to 550 kilograms with a peak near 350 — wildebeest, zebra, buffalo, gemsbok — and the same preference emerges at sites with quite different prey communities. That consistency is what makes it a property of the predator rather than of local availability.
The hyena relationship is the other correction, and it has been known since the 1960s without ever reaching the popular account. Lions scavenge heavily, including taking carcasses from hyenas, which hunt for themselves and do so successfully. Carcasses move in both directions, and the direction is settled largely by numbers: enough hyenas will displace a lioness, and a male coalition will displace any number of hyenas. The noble-hunter-and-thieving-scavenger framing came from documentaries rather than from field records.
Lions hunt mostly at night and in low light. Their eyes carry a tapetum lucidum — a reflective layer behind the retina — which roughly doubles the light available to the photoreceptors and is what makes them shine in a spotlight.
Words used here
Tapetum lucidum
A reflective layer behind the retina that passes light through the photoreceptors twice. Common in nocturnal mammals, and the reason for eyeshine.
Kleptoparasitism
Taking food another animal has caught. Lions and hyenas do it to each other, in both directions, depending on relative numbers.
They count the voices, compare with their own numbers, and decide whether to approach
Established
Specialists would state this without hedging. Multiple independent lines of evidence agree.
Playback of recorded roars from unfamiliar females to wild prides shows that lionesses approach the speaker when they outnumber the simulated intruders and behave cautiously when they do not. The decision tracks the ratio between the number of defenders present and the number of distinct voices heard, rather than the absolute number of either, and changes when nearby pride members provide reinforcement.
Who this applies to
female lions in the Serengeti, tested by playback
Studied in
Panthera leo
You may have heard
“A lion roars to intimidate rivals and can be heard five miles away”
The roar carries a long way — around eight kilometres in good conditions, which is roughly where the five-mile figure comes from — and its function is closer to a census than to intimidation. Playback experiments show lionesses deciding whether to approach based on how many distinct voices they hear against how many companions they have. It is advertisement of numbers between groups that mostly avoid fighting, and the decision is a comparison rather than a reaction.
Why we rate it this way, and what the caveats are
EstablishedHigh confidence
A controlled field experiment in which the number of simulated intruders is set by the experimenter and cannot be confounded with anything the intruders actually do, since they do not exist.
Caveats
Playback presents an intruder that never materialises, so escalation beyond approach cannot be observed.
Contests between males were not tested by this design.
One population; pride composition and density vary elsewhere.
Still unanswered
Do lions assess opponent numbers by counting distinct voices, or by some aggregate property of the chorus that correlates with number?
Approach decisions tracking the ratio of defenders present to voices heard, across one-voice and three-voice playbacks.
How we know
Playing lions the sound of intruders who are not there
Can a lioness tell how many strangers she is hearing, and does she compare it with her own numbers?
Watching real contests cannot answer this, because the two sides differ in far more than number — condition, experience, how badly each wants the ground. A speaker can be controlled precisely. Recorded roars of unfamiliar females were played to wild prides from a concealed loudspeaker, with the number of distinct voices set by the experimenters at one or three, while the number of defending lionesses present was known from individual identification. What the lionesses did next — approach, or hold back — was scored. Because the intruders do not exist, nothing about them can vary except how many voices there are.
What happened
Lionesses approached when they outnumbered the voices and were markedly more cautious when they did not. The decision tracked the *ratio* of defenders present to voices heard rather than the absolute number of either, and changed when nearby pride members offered reinforcement.
What it shows
Numerical assessment in a wild animal, cleanly. The lionesses are not responding to loudness — three distant voices can be quieter than one nearby — but to how many separate animals they hear, weighed against how many they have. It also explains what roaring is for: advertisement of numbers between groups that would mostly rather not fight.
What it does not show
The intruder never materialises, so what happens beyond approach — whether a fight follows, and who wins — cannot be observed by this design. Male contests were not tested. And it does not distinguish counting distinct voices from responding to some aggregate property of the chorus that happens to correlate with number.
The controls — what makes this evidence rather than a story
Number of simulated intruders set by the experimenters, so it cannot correlate with anything else about them.
Playback amplitude controlled, so three voices are not simply louder than one — the cue must be voice count rather than volume.
Defending group size known from individual identification rather than estimated.
Unfamiliar individuals used, so the response is not to a known neighbour.
Trials separated in time and location to avoid habituation to the speaker.
A lion roar carries around eight kilometres in good conditions, which is where the familiar five-mile figure comes from. Its function turns out to be closer to a census than to intimidation. Playing recorded roars of unfamiliar females to wild prides — one voice or three, from a hidden speaker — showed lionesses approaching when they outnumbered the voices and holding back when they did not.
The decision tracked the ratio of defenders present to voices heard rather than the absolute number of either, and it changed when nearby pride members offered reinforcement. Because the intruders do not exist, nothing about them could vary except how many there were — which is what makes this one of the cleanest numerical-assessment experiments in any wild animal.
The mane advertises condition, and its cost is heat
Well supported
Good evidence backs this, though some details remain open.
Presentation of life-sized model lions differing only in mane length and colour showed that females preferentially approached dark-maned models while males avoided them. Long-term records associate dark manes with higher testosterone, better nutrition, longer reproductive tenure and improved offspring survival, and associate mane length with recent fighting success. Manes impose a thermoregulatory cost, and males in hotter regions have shorter and lighter manes, including effectively maneless populations in some areas.
Who this applies to
African lions, tested in the Serengeti and Ngorongoro
Studied in
Panthera leo
You may have heard
“The mane protects a lion's neck in fights”
The protection idea is old and poorly supported; the signalling evidence is strong. Dark manes attract females and deter rival males, and they go with better nutrition, higher testosterone and longer tenure. The cost that keeps the signal honest is heat: a dark mane in a hot climate is expensive to carry, which is why males in the warmest regions have short pale manes or effectively none. Maneless lions are not a separate kind of lion, they are the same trade-off resolved the other way.
Why we rate it this way, and what the caveats are
Well supportedHigh confidence
The model experiment isolates mane characteristics from everything else about a lion and produces opposite, predicted responses in the two sexes; the condition association comes from decades of individual records.
How far it can be extended
The thermal constraint predicts the observed geographic pattern of mane variation, which holds across African populations, though the experimental work is from two.
Caveats
Models cannot reproduce scent or movement, so responses may understate real assessment.
The mane–condition relationship within the long-term dataset is correlational.
Mane variation outside the studied populations is described rather than experimentally tested.
Still unanswered
How much of the geographic variation in manes is a direct thermal response and how much is genetic difference between populations?
Model presentations isolating mane length and colour, the condition and tenure associations, and the thermal cost explaining geographic variation.
How we know
Four dummy lions, identical except for the hair
What is the mane for, and what do other lions read from it?
Comparing real lions cannot isolate the mane, because a male with a magnificent mane differs from one without in size, age, condition and history. The fix was to build the lions. Life-sized models were made in four versions differing only in mane length and mane colour — long dark, long blond, short dark, short blond — and presented to wild lions in the field. Male and female responses were recorded separately, since sexual selection predicts they should respond in opposite directions. Those results were then set against decades of records relating real males' manes to their condition, injuries, tenure and offspring survival, and against the temperature of the regions they lived in.
What happened
Females preferred dark-maned models; males avoided them. In the long-term records, dark-maned males had higher testosterone, better nutrition, longer tenure and better-surviving offspring, while mane length tracked recent fighting success. Manes carry a thermal cost, and males in hotter regions have shorter, lighter manes.
What it shows
That the mane is a signal of condition rather than armour, and that it is honest because it is expensive — carrying a dark mane in the heat costs something only a male in good condition can afford. It also explains the maneless lions of some hot regions, which are not a separate kind of lion but the same trade-off resolved differently.
What it does not show
A model has no scent and does not move, so real assessment may use more than the experiment could offer, and responses may understate it. The mane–condition relationship in the long-term data is correlational. And the experimental work is from two populations, so mane variation elsewhere is described rather than tested.
The controls — what makes this evidence rather than a story
Models identical in every respect except mane length and colour, which is the isolation a field comparison cannot achieve.
Both variables crossed, separating what length signals from what colour signals.
Male and female responses scored separately, testing the opposite predictions sexual selection makes.
Model position and presentation order varied across trials.
Long-term individual records used independently, so the field experiment and the condition association are not the same evidence twice.
The mane was long assumed to protect the neck in fights, and the evidence supports something else. Presenting wild lions with life-sized models identical except for mane length and colour produced opposite responses in the two sexes — females approached dark-maned models, males avoided them — which is the signature of a sexually selected signal rather than armour.
The long-term records agree: dark-maned males had higher testosterone, better nutrition, longer tenure and better-surviving offspring, while mane length tracked recent fighting success. And what keeps the signal honest is temperature. A dark mane is thermally expensive, so only a male in good condition can carry one — which is why males in the hottest regions have short pale manes or effectively none. The maneless lions of Tsavo are not a separate kind of lion; they are the same trade-off resolved the other way.
Words used here
Honest signal
A signal that cannot be faked because producing it is costly in proportion to the quality it advertises. The mane's cost is heat.
Twenty hours of inactivity is not twenty hours of sleep
Well supported
Good evidence backs this, though some details remain open.
Activity budgets from continuous field observation record lions as inactive for approximately twenty hours per day. Inactivity as scored includes resting, grooming, social contact and alert vigilance as well as sleep, and no electrophysiological measurement of lion sleep in the wild exists. The inactivity is consistent with the energetics of a large predator taking infrequent large meals, in which hunting is metabolically expensive and rest conserves the return.
Who this applies to
lions in the Serengeti, from observational activity budgets
Studied in
Panthera leo
You may have heard
“Lions sleep for twenty hours a day”
The figure is real and the word is wrong. What was recorded is *inactivity*, which includes resting, grooming, socialising and lying alert as well as sleeping — and nobody has measured lion sleep in the wild, because doing so needs electrodes on a wild lion. The reason for all that lying about is sound: hunting is expensive, meals are large and infrequent, and resting is how a big predator holds on to what it caught.
Why we rate it this way, and what the caveats are
Well supportedModerate confidence
The inactivity figure is directly and repeatedly observed. Its interpretation as sleep is not supported by any measurement, and this claim is careful to assert only what was scored.
How far it can be extended
Comparable inactivity is reported across lion study sites, though activity budgets vary with prey density and temperature.
Caveats
No electrophysiological measurement of sleep exists for wild lions, so the distinction between rest and sleep is unresolved.
Activity budgets vary with prey density, temperature and season.
Observation is biased towards daylight, when lions are least active.
Still unanswered
How much of a lion's inactive time is actually sleep, and does it show the sleep architecture of other carnivores?
The foraging energetics within which prolonged rest between infrequent large meals makes sense.
The twenty-hour figure is real and the word attached to it is not. What field observation records is *inactivity*, which covers resting, grooming, social contact and lying alert as well as sleeping. Nobody has measured sleep in a wild lion, because doing so would require electrodes on one.
The reason for all that lying about is nonetheless sound. Hunting is metabolically expensive and frequently fails; meals are large and infrequent; and for a large predator, resting is how you keep what you caught. Activity concentrates around dusk and through the night, which is also when observation is hardest — so daylight observation systematically over-samples the inactive part of the day.
Lions are collapsing in West Africa and increasing in fenced southern reserves
Well supported
Good evidence backs this, though some details remain open.
Analysis of trend data from dozens of African lion populations projects continued decline across most of the continent, with the steepest declines in West and Central Africa, while populations in several southern African countries with intensively managed — largely fenced — reserves are stable or increasing. Lions occupy a small fraction of their historical range. A separate small Asiatic population persists in India, differing in social structure in that males associate with females only around mating.
Who this applies to
African lion populations by region, with the Asiatic population noted separately
Studied in
Panthera leo
You may have heard
“Lions live in Africa and their numbers are falling”
Both halves need qualifying. A separate population of Asiatic lions survives in Gir in India, and it is not a remnant of the African species in any casual sense — it has its own social structure, in which males and females largely live apart. And a single continental trend hides the finding: lions are collapsing across West Africa while increasing in some intensively managed southern reserves. Averaging those produces a number that describes nowhere.
Why we rate it this way, and what the caveats are
Well supportedModerate confidence
The regional divergence is robust and is the substantive result. Absolute figures carry wide uncertainty because census quality varies enormously between populations, particularly in the regions declining fastest.
How far it can be extended
Trends were compiled from populations across the range; the divergence between regions is the finding rather than an artefact of sampling.
Caveats
Trend data quality varies enormously between populations and regions.
Projections carry wide uncertainty where the underlying census is weak.
Whether fenced, intensively managed populations are the desired model at scale is a value judgement rather than an ecological result.
Still unanswered
Can unfenced lion populations be stabilised outside intensively managed reserves, or is coexistence at scale unsolved?
Bauer et al., 2015 · Proceedings of the National Academy of Sciences
Regional trend projections and the association between stability and intensive management.
A single figure for lion numbers describes nowhere. Populations are collapsing across West and Central Africa while several intensively managed — largely fenced — reserves in southern Africa hold stable or increasing populations. Averaging those produces a number that matches no actual place, and the divergence is the finding.
Lions occupy a small fraction of their historical range. The pressures are habitat loss, loss of prey to bushmeat hunting, and conflict with livestock keepers, which is the one that matters most in practice — a lion that takes cattle is usually killed, and from the herder's position that is a rational response to losing an asset.
The Asiatic lions of Gir are a separate case. Several hundred animals in a single population, recovered from a few dozen a century ago, and differing socially: males associate with females mainly around mating rather than living with the pride. A single population of that size is vulnerable to one disease outbreak, which is why establishing a second one has been discussed for decades.
Conservation
What the fenced-reserve result does and does not settle
That intensively managed reserves hold stable lion populations is an ecological result. Whether fenced, closely managed populations are the outcome anyone wants at scale is a value judgement, and the evidence does not settle it — fencing changes dispersal, gene flow and the ecology of everything else inside. NatureHQ reports the trend and the association with management, and does not have a position on the policy question.
Where this applies: Global principle; conservation policy is set nationally by range-state authorities.
Does the cooperative-hunting explanation fail everywhere, or only where prey is abundant?
Why it matters: The measurement that overturned it comes from the Serengeti, which is unusually prey-rich. Where prey is scarce, feeding benefits of cooperation could be real — and if so, sociality would have different explanations in different places.
What would settle it: Per-capita intake measured against group size in an arid, prey-poor system with individually known lions.
Why do unrelated males accept each other into coalitions?
Why it matters: Kin selection explains brothers cooperating. Unrelated partners share reproduction with a non-relative, and mutual benefit explains why it can pay — but not how partners are chosen or how cheating is constrained.
What would settle it: Paternity assignment within unrelated coalitions across many takeovers, showing how reproduction is actually divided.
How much of a lion's food is scavenged rather than hunted?
Why it matters: Scavenging is known to be substantial and its proportion is not established, which leaves a basic question about lion ecology open — and it is the number the hyena comparison turns on.
What would settle it: Continuous tracking with accelerometry and audio at carcasses, distinguishing kills from arrivals at existing ones.
How much of a lion's inactive time is actually sleep?
Why it matters: The twenty-hour figure is quoted everywhere as sleep and was recorded as inactivity. Nobody knows the lion's sleep architecture, or whether it resembles other carnivores.
What would settle it: Electrophysiological recording from wild lions, which has been achieved in other large mammals and not yet in these.
NatureHQ publishes its own gaps. This record is at 90% completeness against what we would call a finished subject.
no research from the last few years is attached — check for newer work
Asiatic lions are covered in outline; their differing social structure deserves its own treatment and the literature on them is substantial.
Almost all the social biology here comes from the Serengeti and Kruger, and lion behaviour in arid and West African systems is far less studied.
Human–lion conflict is named as the dominant conservation pressure and not treated properly, because doing so responsibly needs the social science alongside the ecology.
Lion communication beyond roaring — scent marking, contact calls, facial and tail signals — is essentially absent here.
Last reviewed 2026-08-11 · 9 claims · 55 search questions answered on this page