Dunbar's number: how many close friends can you actually have?
Robin Dunbar's research gives a surprisingly precise answer to the most important question about your social circle — and the consequences change everything.
Most adults can maintain around 150 stable relationships — Dunbar’s number — organised into concentric layers of 5, 15, 50, and 150 close friends and acquaintances. The exact count matters less than the distribution: research by Dunbar (1992) and Hill & Dunbar (2003) shows it is the inner layers, not the outer ones, that predict how supported and connected you actually feel.
What Dunbar found
How many friends can you really have?
Not in the LinkedIn sense. Not “people whose names you recognise”. Actual stable relationships — people whose lives you broadly know about, whose move you’d help with, the ones you’d call when something goes sideways.
Robin Dunbar, a British anthropologist at Oxford, has a startlingly precise answer. He doesn’t call it this; others did. Dunbar’s number. Around 150.
He didn’t get there in 1992 with a survey. He got there by measuring monkeys [1]. In the study that now sits in every anthropology textbook, he correlated the relative size of the neocortex — the outer layer of the brain — with the average group size of different primate species. The correlation was remarkably clean: the bigger the neocortex, the bigger the stable group an individual can live inside.
When he projected the curve up to the human neocortex, the answer that fell out wasn’t one anyone was expecting: roughly 148. Dunbar himself rounded to 150.
What started as a theoretical prediction turned out to match the world surprisingly well. The typical size of historical farming villages: around 150. The headcount of a Roman centuria, or a modern military company: about 150. The average number of names in pre-digital address books that researchers studied: 150, give or take. The number of people you’d send a wedding invitation to before parents and in-laws start padding the list — also, roughly, in this region.
It’s not that 150 is a magic number. It’s an estimate of the upper bound on stable, simultaneously-maintained relationships a human brain can carry. You can have more relationships than that. You can’t have more active stable ones.
The five layers
What makes the original 150 more interesting than just a point on a scale is later work by Russell Hill and Robin Dunbar, in 2003 [2]. It showed that the same correlation explaining the 150 also explains what happens inside it. Relationships aren’t all the same depth. They sort themselves into concentric layers — and each layer is, consistently, about three times the size of the one inside it.
5 people — the inner circle. A partner, a best friend, maybe a sibling, someone from childhood who never quite drifted away. These five are the ones you call when something falls apart at two in the morning. They’re rare because they’re expensive — they cost time, attention, and honest vulnerability. How you express and receive care within this layer shapes who stays — if you haven’t thought about this before, the research on friendship languages is worth the detour.
15 people — what Dunbar calls the “sympathy group”. Close friendships. People whose grief would actually shake you. A full layer here is a quiet privilege; many adults have fewer than 15 here without realising it.
50 people — the “band”. An extended close circle. People you’re glad to bump into, who you eat or celebrate with on a regular cadence, whose birthday you remember without your calendar prompting you.
150 people — your stable relationships overall. The former colleagues, the old university friends, the in-laws of in-laws, the neighbours you missed when they moved. You’d show up at the funeral of any of them. For most, you broadly know what they do for a living.
500 people — acquaintances. You recognise them, can match a name to the face, broadly recall the context where you first met. But a phone call would be a surprise.
What these layers share is an awkward arithmetic: each inner layer needs disproportionately more attention per person. The five closest people might consume forty percent of your social bandwidth; the next 145 split the rest. This isn’t a moral judgement — it’s just how attention distributes. And it’s strikingly consistent across cultures and centuries.
Friendships, real friendships, take time to build and maintenance to sustain. There are no shortcuts.
— Robin Dunbar, How Many Friends Does One Person Need? (2010)
Try it on yourself
Theory is one thing. Looking at your own layers is another.
We built a small calculator that asks you how many people are currently in each of your circles, and shows you where you sit relative to Dunbar’s estimates. It saves nothing, sends nothing to a server. It runs in your browser, closes, and is gone.
Most people are surprised. Sometimes by how crowded the outer layers are — how many half-tended acquaintances have piled up over the years. Sometimes by how sparse the inner one is. Both findings are useful. They’re the start of an honest answer to the question of who is actually in your life right now.
What most people get wrong
Dunbar’s number is one of the most widely cited and widely misunderstood research findings of the last thirty years. Three misreadings keep coming up.
First: “Social media has made this irrelevant.” This is the most common reaction, and it was once a fair hypothesis. Dunbar tested it himself in 2016 [3], in a study looking at behaviour on Facebook and Twitter. The result: the number of relationships people actually maintain — not “follow” but interact with, visit, call — sits in the classic Dunbar range. Platforms multiplied reach. They didn’t lift capacity. Someone with 800 Facebook friends is still a person with about 5/15/50/150 real relationships, plus 650 pictures with names attached.
Second: “150 is a hard ceiling.” It isn’t. It’s a soft maximum that drifts with maintenance. An old friend you haven’t seen in years slips from your 50 layer into your 150, then into the 500. A colleague you work with daily can climb into your 15 within months. The numbers describe capacity under normal upkeep. Neglect contracts them inward; intense seasons can stretch them temporarily.
Third: “I can be close to everyone.” The arithmetic says no. If your inner layer holds five people and you give each, say, one hour of real attention per week — five hours total — and tried to spread that same time across fifty people, each one would get six minutes. Six minutes doesn’t build closeness. Bandwidth per person scales inversely with circle size. None of which means you have to be cruel. It means the choice of who’s in your inner circle gets made anyway — either deliberately, or by drift.
What Dunbar's research says
Around 150 stable relationships, organised in layers of ~5/15/50/150. Maintenance is a precondition — the numbers shrink when attention disappears.
What people often believe
“I have 800 friends on LinkedIn.” — Reach isn’t capacity. Most of those 800 sit in the outer 500 layer, or have already drifted out of your active circle.
What does this mean in practice?
The lesson from Dunbar isn’t “have fewer friends”. It’s a harder one: be honest about which 150 you’ve chosen — and which 5, 15, 50.
Most people never actively choose. Their inner layers fill by inertia: who happens to be geographically close, who was in school with them, who sits next to them at work. None of that is wrong — but it isn’t the result of a decision either. It’s the result of probability. And probability is a poor algorithm for picking the people who’ll walk through your life with you.
Three concrete thoughts, if you’re now wondering what to do with any of this:
First: do an honest audit. Who is actually in your 5? Who would you call at 2 a.m.? Who would you call with good news first? If those two lists don’t overlap, that’s information.
Second: decide deliberately which friendships to deepen. It can’t be all of them. But it can be one, this year. One person currently in your 50 who really belongs in your 15. Who is that?
Third: make maintenance routine, not heroic. The biggest damage in adult relationships rarely comes from arguments. It comes from radio silence that creeps in — and then, eventually, feels too awkward to break. A message every few weeks, a call on a birthday, a lunch on the calendar every few months: that’s what holds a layer in place. It’s trivial, which is exactly why it’s so easy to forget. If you are unsure how often to reach out, our piece on how often to check in with friends works through the contact cadences that match each layer.
We build Endearist for exactly this reason. Because a small nudge at the right moment — “you haven’t messaged Sarah in three months” — isn’t paternalism. It’s what keeps your 50 layer from quietly becoming your 150. But that’s another story.
What matters here: the 150 will get chosen one way or another. The only question is whether you choose them, or whether you let it happen by chance.
References
-
Reference Dunbar, R. I. M. (1992). Neocortex size as a constraint on group size in primates. Journal of Human Evolution, 22(6).
https://doi.org/10.1016/0047-2484(92)90081-J -
Reference Hill, R. A., & Dunbar, R. I. M. (2003). Social network size in humans. Human Nature, 14(1).
https://doi.org/10.1007/s12110-003-1016-y -
Reference Dunbar, R. I. M. (2016). Do online social media cut through the constraints that limit the size of offline social networks? Royal Society Open Science, 3(1).
https://doi.org/10.1098/rsos.150292
FAQ
What is Dunbar's number?
Dunbar's number describes the maximum number of stable social relationships a person can maintain at once — about 150. It comes from Robin Dunbar's research on the neocortex of primates, extrapolated to humans.
Why exactly 150?
Dunbar derived the figure from the correlation between neocortex size and average group size in primates. Extrapolated to humans, the curve predicts ~150 — a number that recurs in historical village sizes, military companies, and pre-digital address books.
Does the number still hold today, with social media?
Studies show that despite thousands of followers, the number of relationships people actually maintain stays in the classic Dunbar range. Social media expands reach, not capacity.
Can I maintain more than 150 relationships if I really try?
Research suggests discipline shifts the upper end modestly, not radically. People who actively maintain 200 or 250 relationships usually trade depth in their inner layers for breadth. The question is less 'how many can you' than 'which 150 should they be'.
How many close friends should an adult have?
**Dunbar's research** points to around **5** people in your innermost circle and **15** in your sympathy group — these are the close friends whose loss would genuinely shake you. Most adults land below both figures. **Hill & Dunbar (2003)** showed these layers are consistent across cultures, not aspirational targets. If you have **two or three** close friends you can call at 2 a.m., you are not failing; you are statistically typical. The useful question is whether those relationships are actually maintained, not whether the count matches a diagram.
Is it normal to have only one or two close friends?
Yes — and research backs it. **Dunbar (1992)** placed the inner circle at roughly **5** people, but that is a capacity ceiling, not a floor. Survey data consistently finds many adults reporting **fewer than three** confidants. One or two close friendships, maintained well, deliver most of what the research attributes to a full inner circle: reduced stress, longer life expectancy, resilience during hardship. The risk of a very small inner circle is fragility: if one friendship breaks or a person moves away, the whole layer collapses. Keeping even **one backup friendship** in your **15-layer** matters.
Can you have too many friends?
Not in the abstract — but you can have more active friendships than your attention budget can sustain. **Dunbar (1992)** showed that the neocortex imposes a ceiling of roughly **150** stable relationships; beyond that, the brain cannot track the reciprocal obligations that define a real relationship. In practice, the danger is subtler: spreading attention across too many people in your **50-layer** starves the **5** and **15** layers of the time they need. **Hill & Dunbar (2003)** found that inner-circle relationships require disproportionate upkeep — roughly 40 % of social bandwidth for just **5** people. More acquaintances rarely hurts. More people you call close friends, without the hours to match, produces a social circle that feels full and functions empty.
Why is it harder to make friends after 30?
Three structural forces compound after 30. First, the environments that produced effortless proximity — shared dormitories, seminars, sports teams — disappear. **Dunbar (2010)** notes that repeated unplanned contact is the most reliable engine of closeness, and adult life systematically removes it. Second, existing slots in your **15** and **50** layers are already occupied; a new person has to displace someone rather than join an empty seat. Third, the investment threshold rises: **78 hours** of contact over roughly three years is the estimate researchers associate with moving from acquaintance to close friend, and that is time most working adults genuinely do not have available. None of this is permanent. It means you have to be deliberate — schedule the lunch, join the recurring thing, show up consistently.
What counts as a 'friend' vs. an acquaintance?
**Dunbar's layers** offer a useful operational distinction. An acquaintance (**500-layer**) is someone you recognise and can place in context — you know their name and broadly where you met. A friend (**50-layer**) is someone whose birthday you remember unprompted and whose move you would help with. A close friend (**15-layer**) is someone whose grief would affect you and who knows the current version of your life, not just its history. The line between acquaintance and friend is usually _shared time under low-stakes conditions_; the line between friend and close friend is usually _shared vulnerability_. **Hill & Dunbar (2003)** found these distinctions map onto measurable differences in contact frequency and emotional investment, not just intuition.
How many friends do most people lose per decade?
**Dunbar (2010)** estimated that the average person loses roughly **half** their active friendships each **seven-year** period if no deliberate maintenance is applied. The outer layers — your **50** and **150** — erode fastest, because those relationships run on proximity and habit. When a job changes or a city changes, the structural supports vanish and contact simply stops. The inner **5** are more durable but not immune: research on long-distance friendships suggests that even close friendships require _some_ contact to stay in the inner layer rather than drifting outward. The antidote is low-effort, high-frequency contact — not grand reunions but regular small signals. A short message every few weeks can hold a friendship in your **15** that would otherwise slide to your **150** within a year.
Are online friends real friends?
**Dunbar (2016)** tested this directly, comparing online interaction patterns to offline relationship quality on Facebook and Twitter. His finding: online contact can _maintain_ an existing close friendship but rarely _builds_ one from scratch to inner-circle depth. The mechanism seems to be that closeness depends on cues — tone, timing, physical presence — that text-based platforms strip out. Online friends who also meet in person, even occasionally, can absolutely occupy your **15** or even your **5**. Purely online relationships tend to cluster in the **50** or outer layers. That is not a moral verdict; it is a description of what the channel supports. If an online friendship matters to you, the research suggests the investment that deepens it is finding a way to meet.
How do I know if I'm in someone's inner circle?
**Hill & Dunbar (2003)** found that inner-circle membership (**the 5-layer**) correlates with three observable patterns: _contact frequency_ (you hear from them without needing to prompt), _initiation symmetry_ (both sides reach out, not just one), and _crisis availability_ (they respond quickly when something goes wrong, not just when it is convenient). A fourth signal, less measurable but reliable: they know things about your current life — your job situation, your stress, your plans — not just your history. If you are unsure whether you are in someone's **5 or 15**, the honest test is whether they would call _you_ with bad news before most people. That is the real inner circle: not warmth, not history, but _mutual first-call status_.