Skip to content
Fraily

The Science of Friendship

The Science of Friendship: What Biology and Evolution Reveal

Friendship is not merely a cultural phenomenon. It has deep biological roots — governed by five biochemical systems, limited by the cognitive capacity of our brain, and documented in species from chimpanzees to dolphins. This overview summarizes what neurobiology, evolutionary research, and comparative behavioral science have discovered about the mechanisms behind friendship.

By Fraily EditorialReading time approx. 18 minutes

What Does Science Say About Friendship?

Friendships are governed by an interplay of multiple neurotransmitters, neuropeptides, and hormonal systems that are remarkably conserved across species. The same or homologous neural circuits control social behaviors in birds, rodents, primates, and humans (Brent et al., 2014).

This cross-species conservation argues against the assumption that friendship is a cultural construct. It points to a shared evolutionary origin. The Social Brain Hypothesis (Dunbar, 1998) holds that group living created selection pressure for larger and more complex brains — a finding consistent across species.

Moreover, the tendency toward social bonds — the affiliative tendency — has a heritable basis. This has been demonstrated in humans, yellow-bellied marmots, and rhesus macaques. Sociality is thus a trait on which natural selection can act.

Five Biochemical Systems

Five systems are at the center of current research on the neurobiology of friendship. Each serves a different function.

SystemFunctionKey Detail
OxytocinTrust, prosocial decisionsOnly active with existing friends
β-EndorphinCement bonds, generate warmth20–100x more potent than morphine
DopamineSocial memories, preferencesVia the ventral tegmental area
SerotoninSocial perception and responseGenetic variants influence integration
HPA axisStress regulationFriends dampen cortisol response

After Brent et al. (2014).

Oxytocin and Trust

Oxytocin was long associated only with mother-child bonding and romantic love. But a key study on wild chimpanzees (Crockford et al., 2013) changed this picture: oxytocin levels rose after mutual grooming — but only between existing friends, not with unfamiliar group members.

Oxytocin thus does not facilitate social interaction in general but is specifically involved in maintaining existing reciprocal bonds. For everyday life this means: the neurobiological effect of interaction is stronger with friends than with strangers. Every meeting with an existing friend strengthens the bond at a biochemical level.

Endorphins as Bonding Agents

β-Endorphins are the central neurochemical mechanism behind the formation and maintenance of friendships. While oxytocin facilitates entry into social interaction, endorphins cement the actual bond (Machin & Dunbar, 2011).

Their activation creates a feeling of warmth, calm, relaxation, trust, and connectedness. The health effects work along two pathways: they directly lift mood and act as a natural antidepressant, and they activate the immune system, particularly natural killer cells (Dunbar, 2025). More in our article on endorphins and friendship.

A remarkable finding: a study on group rowing showed that participants released more endorphins when rowing together than when rowing alone — at the same level of physical exertion. The social component amplifies the biochemical effect. Another study found that individual baseline pain tolerance correlates with the size of one’s close friend circle.

Dunbar’s Number

The cognitive capacity of our brain limits the number of stable social contacts to about 150— the famous Dunbar’s Number. Within this network, relationships are organized in concentric layers, each requiring a different level of emotional investment. More in our article on Dunbar’s Number.

The innermost layer comprises 1–2 closest confidants, followed by 5 close friends (weekly contact), 15 good friends, and 50 acquaintances. In the close 5-layer, only one person is replaced per decade. In the outer layers, up to 30% change annually.

The key insight for everyday life: your closest five friends need at least weekly contact. Fall below that and emotional closeness begins to measurably decline within a few months. The form of contact does not matter — phone call, text message, or meeting in person are equally effective.

Friendship in Animals

Friendship is not a purely human phenomenon. Brent et al. (2014) define friendship as bidirectional, affiliative interactions whose frequency and consistency distinguish them from non-friendships. This definition deliberately avoids assumptions about feelings or motives and enables cross-species comparisons.

Friendship has been documented in chimpanzees, baboons, dolphins, elephants, hyenas, and even birds. The drivers of group formation differ across species: in primates it was likely the shift from nocturnal to diurnal living and the resulting predation pressure; in carnivores, cooperative hunting; in birds, cooperative brood care.

The key finding: animals with closer social networks show lower baseline cortisol levels and higher reproductive success. Friendship thus has measurable adaptive advantages not only in humans but also in animals.

Friends Reduce Stress — Biochemically

The HPA axis (hypothalamic-pituitary-adrenal axis) connects stress regulation with social behavior. Animals and humans with closer social networks show lower baseline cortisol levels, and the presence of friends dampens the stress response. More in our article on friends and stress reduction.

An important distinction: stress reduction is a proximate mechanism, not the evolutionary function of friendship. The ultimate reasons lie in increased survival and reproductive success. But the proximate effect is what you feel: after a conversation with a close friend, you feel calmer, safer, more resilient.

Limits of the Research

The precise molecular processes through which serotonin influences social behavior remain poorly understood. Many findings on oxytocin and endorphins come from animal or laboratory studies and cannot be straightforwardly transferred to everyday human life.

The Social Brain Hypothesis is correlational — larger brains could also be explained by other factors. And transferring animal findings to human friendship is methodologically demanding, as human friendships have additional cultural and cognitive dimensions.

What the research clearly shows, however: friendship is neither a luxury nor a cultural construct. It is a biologically anchored, evolutionarily developed trait with measurable effects on the brain, the body, and survival.

Put the science to work

Your brain is built to nurture friendships. Fraily helps you do just that — with a FriendshipValue that shows which relationships are alive and where you might want to reach out again.

Frequently Asked Questions

What happens in the brain during friendship?
Five biochemical systems govern friendships: oxytocin (trust), β-endorphins (bonding), dopamine (social memories), serotonin (social perception), and the HPA axis (stress regulation). Endorphins play an especially central role — they are 20–100 times more potent than morphine and create a feeling of warmth and connectedness.
Is friendship innate or learned?
Both. The tendency toward social bonds (affiliative tendency) has a heritable basis — demonstrated in humans, marmots, and rhesus macaques. But which specific friendships form and how they are lived is strongly shaped by environment and culture.
How many friends can a person have?
Dunbar’s Number limits stable social contacts to about 150. Of those, roughly 5 are close friends (needing weekly contact), 15 good friends, 50 acquaintances. Each layer requires a different level of time and emotional investment.
Does friendship exist in animals?
Yes. Friendship has been documented in chimpanzees, dolphins, elephants, hyenas, and even birds. Brent et al. (2014) define friendship as bidirectional, affiliative interactions whose frequency and consistency distinguish them from non-friendships — deliberately avoiding assumptions about feelings.

Sources

  1. Brent, L. J. N., Chang, S. W. C., Gariépy, J.-F. & Platt, M. L. (2014). The neuroethology of friendship. Annals of the New York Academy of Sciences, 1316, 1–17.
  2. Dunbar, R. I. M. (2025). Why friendship and loneliness affect our health. Annals of the New York Academy of Sciences, 1545, 52–65.
  3. Machin, A. J. & Dunbar, R. I. M. (2011). The brain opioid theory of social attachment: A review of the evidence. Behaviour, 148, 985–1025.
  4. Crockford, C. et al. (2013). Urinary oxytocin and social bonding in related and unrelated wild chimpanzees. Proceedings of the Royal Society B, 280, 20122765.
  5. Pearce, E., Wlodarski, R., Machin, A. & Dunbar, R. I. M. (2017). Variation in the β-endorphin, oxytocin, and dopamine receptor genes is associated with different dimensions of human sociality. PNAS, 114, 5300–5305.
  6. Keverne, E. B., Martensz, N. D. & Tuite, B. (1989). Beta-endorphin concentrations in cerebrospinal fluid of monkeys are influenced by grooming relationships. Psychoneuroendocrinology, 14, 155–161.

Alle Artikel in diesem Themenbereich