How Many Friends Do We Need, and How Long Does It Take to Make One?

Count the people you would genuinely call friends. Not contacts, not the names in your phone - the people you would call first with good news. Most people find the number is smaller than they expected, and many quietly wonder whether it ought to be bigger.

Two pieces of research are usually quoted in reply. One says humans can maintain about 150 relationships. The other says it takes about 200 hours to turn a stranger into a close friend. Both come from real studies. Both are more interesting, and more uncertain, than the headlines suggest.

Where the number 150 comes from

In 1992 the anthropologist Robin Dunbar published a study of primates showing that species with a larger neocortex, relative to the rest of the brain, tend to live in larger social groups. His explanation, developed in a 1998 review into what he called the 'social brain hypothesis', was that keeping track of relationships - who is allied with whom, who owes whom a favour - is mentally demanding, and that brain capacity puts a ceiling on how many relationships an animal can manage.

Extend the primate line to the size of the human neocortex and you get a predicted group size of about 150. That figure became famous as 'Dunbar's number', and it has since been quoted far beyond anthropology.

Dunbar and his colleague Russell Hill then looked for it in everyday life. In a study published in 2003, they used the exchange of Christmas cards as a way of mapping people's social networks in a Western sample. The maximum network averaged 153.5 people, and 124.9 when counting only those who were contacted directly - remarkably close to the prediction. About 21% of each network were relatives, a proportion that stayed fairly stable.

Circles within circles

The 150 is only the outer ring. In 2005, Wei-Xing Zhou, Didier Sornette, Hill and Dunbar pooled data on how human groups are organised and found not a smooth spread of sizes but a series of preferred layers, each roughly three times larger than the one inside it: approximately 3 to 5 people, then 9 to 15, then 30 to 45, and so on outward.

It is tempting to map this onto ordinary life: a handful of people you would turn to in a crisis, a slightly larger ring of close friends, then good friends, then a wider circle of people who matter to you but whom you see less. The Christmas-card study fits that picture in one respect: how often people were in touch depended strongly on emotional closeness, as well as on practical factors such as distance.

Is 150 a real limit?

Here the honest answer is: not as a precise number. In 2021, Patrik Lindenfors, Andreas Wartel and Johan Lind re-ran the kind of analysis that produced the original estimate, using newer primate data and different statistical methods. Depending on the method, the average predicted human group size came out at 69 to 109, or at 16 to 42.

More telling were the uncertainty ranges. The 95% confidence intervals ran from 4 to 520 in one analysis and from 2 to 336 in the other. With error bars that wide, the authors concluded that specifying any one number is futile, and that a cognitive limit on human group size cannot be derived this way.

This does not prove there is no limit. Time is finite, and nobody can maintain an unlimited number of close relationships. What it shows is that the neat figure of 150 is much less certain than its fame suggests, and that the Christmas-card result is one study from one kind of society.

How long it takes to make a friend

The second famous number is about time. In 2019 Jeffrey Hall, a communication researcher at the University of Kansas, published two studies. In the first, 355 adults who had recently relocated estimated how much time they had spent with someone they had met since the move, and how close they now felt. In the second, 112 first-year university students reported the hours they had spent with two new acquaintances at three points over nine weeks.

In both studies, more hours together went with closer friendships. Hall's estimates were roughly 50 hours of time together for a casual friendship, around 90 hours for a friendship, and more than 200 hours for a close one.

How the time was spent mattered too. Time spent on leisure together - hanging out, joking around, doing things for fun - predicted closeness, while time spent side by side at work counted for less. Certain kinds of everyday talk also predicted friendships growing closer between one survey and the next.

Treat these numbers as rough guides, not a recipe. They rely on people's own estimates of past hours, which are fallible. They come from online volunteers and students at one American university. And they are correlational: people spend more time with those they already like, so time together is partly a result of friendship, not only a cause.

What this means for you

• A small inner circle is normal. The layered structure suggests that only a handful of people occupy the innermost ring for almost everyone. Having few very close friends is not in itself a sign that something is wrong.
• Friendship is built from hours, and hours are a choice. If you want to get closer to someone, the most consistent lever in the research is simply more time together, especially relaxed, enjoyable time.
• Do not wait for work to do it for you. In Hall's data, leisure time mattered more than time spent working side by side.
• Be patient in a new place. At a few hours a week, 200 hours takes many months. Feeling lonely after a move is common and does not mean it will stay that way.

The friendship style test on this site looks at how you tend to show up for other people, the need to belong test measures how strongly you want to be part of a group, and the loneliness test gives a snapshot of how connected you feel right now. None of them can tell you the right number of friends, because there is no single right number. If loneliness has been heavy and persistent for a long time, talking to a doctor or a counsellor is a sensible step.