Oxytocin and Trust: The Science of the Trust Molecule

In 2005, a group of volunteers at the University of Zurich sat in a laboratory, tilted their heads back, and sprayed a fine mist into their nostrils. Half received oxytocin – a nine-amino-acid neuropeptide best known for its role in childbirth and breastfeeding. The other half received a saltwater placebo. Neither the participants nor the researchers running the session knew who received what. Then each volunteer was asked to do something deceptively simple: trust a complete stranger with their money.

What happened next became one of the most cited experiments in the history of behavioural economics and launched an entire field of research into oxytocin’s role in human social behaviour. It also popularised a label that would prove both powerful and problematic – the “trust hormone.”

The Landmark Trust Experiments

The study, published in the journal Nature by Michael Kosfeld, Markus Heinrichs, Paul Zak, Urs Fischbacher, and Ernst Fehr, used a well-established behavioural economics paradigm called the trust game. The rules are straightforward: an “investor” is given a sum of money – in this case 12 monetary units – and can transfer any portion to an anonymous “trustee.” Whatever amount is transferred gets tripled by the experimenters, and the trustee then decides how much, if anything, to return. The investor’s decision to send money is a measurable act of trust – they are making themselves vulnerable to another person’s goodwill.

The results were striking. Among the 29 participants who received intranasal oxytocin, 13 invested the maximum amount, trusting a stranger with everything they had. In the placebo group of 29, only 6 showed the same level of trust. On average, the oxytocin group transferred 17% more money to the trustee than the placebo group.

But the researchers didn’t stop there. They ran a crucial control condition – one that would elevate this study above a simple demonstration of risk-taking. In a separate experiment, participants played a version of the game where returns were determined not by a human trustee but by a random number generator. In this “risk game,” oxytocin had no effect whatsoever on investment behaviour. Participants given oxytocin were no more willing to take financial risks when another person wasn’t involved.

This was the key finding. Oxytocin didn’t simply make people reckless with money or more tolerant of risk. It specifically increased willingness to accept social risk – the vulnerability that comes with depending on another human being. The neuroscientist Antonio Damasio, writing a commentary on the study in the same issue of Nature, called the results “remarkable” and suggested they opened a new window into the biological foundations of moral behaviour.

Paul Zak and the Trust Molecule

While the Kosfeld et al. experiment was a five-author collaboration, it was neuroeconomist Paul Zak of Claremont Graduate University who became most closely associated with the oxytocin-trust narrative in the public imagination. Zak had been studying the economics of trust since the early 2000s, and the 2005 Nature paper gave him a biological mechanism to anchor his ideas.

Zak’s subsequent research went far beyond the laboratory. He measured blood oxytocin levels in people at weddings, finding spikes in the hormone among both participants and observers of the ceremony. He conducted “hugging experiments,” demonstrating that physical touch triggered oxytocin release. He showed that watching emotionally engaging narratives – stories about a father and his terminally ill son, for example – caused oxytocin levels to rise, and that higher oxytocin responses predicted greater subsequent generosity toward strangers.

In his 2012 book The Moral Molecule, Zak wove these findings into a grand narrative: oxytocin was the biochemical basis of morality itself, the molecule that enabled empathy, generosity, and the cooperative behaviours that hold societies together. His popular TED talk on the subject – viewed over a million times – cemented the idea in public consciousness. Oxytocin wasn’t just the “love hormone” associated with maternal bonding; it was the “trust molecule,” the chemical key to human cooperation.

Zak’s work was genuinely pioneering. He was among the first researchers to bridge neuroendocrinology and economics, founding the field that became known as neuroeconomics. His studies showed that oxytocin wasn’t simply a reproductive hormone but played a role in a sweeping range of social behaviours. The New York Times columnist Matt Ridley called it “one of the most revealing experiments in the history of economics.”

Oxytocin, Generosity, and Cooperation

The trust game was only the beginning. Researchers quickly turned to other economic games to explore whether oxytocin influenced broader prosocial behaviour – and the early results were encouraging.

In a 2009 study published in the Annals of the New York Academy of Sciences, Jorge Barraza and Paul Zak demonstrated a direct link between empathy, oxytocin, and generosity. Participants watched short video clips – one emotionally engaging (a father speaking about his young son’s terminal cancer), the other neutral. Those who watched the emotional video showed a 47% increase in blood oxytocin levels compared to baseline. Critically, the magnitude of their oxytocin response predicted how generous they were in a subsequent ultimatum game, where they divided real money with a stranger.

Other studies extended these findings. In dictator game experiments – where one player allocates money to another with no obligation to share – intranasal oxytocin increased generosity by up to 80% compared to placebo. Participants given oxytocin were more likely to share their endowment and offered larger portions to anonymous recipients.

Research into oxytocin and cooperation also gained momentum. Studies found that oxytocin enhanced cooperative behaviour in public goods games, where individuals must decide how much to contribute to a shared pool. The hormone appeared to facilitate the kind of prosocial decision-making that economists had long struggled to explain – why people cooperate even when strict self-interest would dictate otherwise.

These findings connected oxytocin to the broader framework of social bonding. The same neuropeptide that strengthened the bond between mother and infant, that facilitated pair-bonding in monogamous prairie voles, seemed to operate on a wider canvas in humans – promoting trust, empathy, generosity, and cooperation across a range of social interactions with both familiar and unfamiliar people.

The Nuanced Picture

As the “trust molecule” narrative gained traction in the media and popular science, a more complicated picture was emerging in research laboratories around the world. The initial excitement gave way to findings that were harder to reconcile with a simple story.

Replication Challenges

In 2015, Gideon Nave, Colin Camerer, and Michael McCullough published a comprehensive critical review in Perspectives on Psychological Science titled “Does Oxytocin Increase Trust in Humans?” Their conclusion was sobering: the original Kosfeld et al. finding “has not replicated well.” The authors examined the evidence across three approaches – intranasal administration studies, plasma oxytocin correlations, and genetic polymorphism analyses – and found that none provided robust, convergent evidence that oxytocin reliably increases human trust.

The review was part of a broader reckoning within psychology and neuroscience. Many early oxytocin studies used small sample sizes – the original trust game study had just 58 participants – raising concerns about statistical power and the reliability of effect sizes. Walum, Waldman, and Young (2016) published an influential analysis in Biological Psychiatry arguing that most intranasal oxytocin studies were severely underpowered, meaning there was a high probability that many published findings did not represent true effects.

Context-Dependency: The Dark Side of Oxytocin

Perhaps more surprising than the replication difficulties was research suggesting that oxytocin’s social effects were highly context-dependent – and not always positive.

In 2009, Simone Shamay-Tsoory and colleagues at the University of Haifa published a striking study in Biological Psychiatry. Using a game-of-chance paradigm, they found that intranasal oxytocin increased envy when participants received less money than another player, and increased schadenfreude (gloating) when they received more. Oxytocin didn’t simply enhance warm, prosocial feelings – it amplified social comparison emotions across the board, including decidedly antisocial ones.

Carsten De Dreu and colleagues at the University of Amsterdam uncovered another troubling finding. In a series of experiments published in the Proceedings of the National Academy of Sciences in 2011, they showed that intranasal oxytocin promoted ethnocentrism – increasing favouritism toward one’s own group while simultaneously increasing derogation of out-group members. Oxytocin made people more trusting and cooperative toward those they perceived as part of their in-group, but it didn’t extend the same generosity to outsiders. In some conditions, it even increased willingness to harm out-group members to benefit in-group members.

These findings led researchers to reframe oxytocin’s social role. Rather than a universal “trust molecule,” oxytocin appeared to function more as a social salience amplifier – increasing the intensity and importance of social cues and contexts, whether those cues pointed toward trust or toward tribal bias, toward empathy or toward envy.

Trust Beyond the Lab

Despite the scientific controversies, the connection between oxytocin and trust has influenced thinking far beyond academic psychology. The research sparked interest across organisational science, negotiation theory, and leadership studies.

Organisations and Leadership

Paul Zak himself extended his research into the corporate world with his 2017 book Trust Factor: The Science of Creating High-Performance Companies. Drawing on his oxytocin research, Zak argued that organisational cultures fostering trust – through autonomy, recognition, and transparency – produced measurable neurochemical responses associated with engagement and performance. His framework suggested that management practices promoting interpersonal connection effectively leveraged the same oxytocin-mediated circuits identified in laboratory trust games.

While the direct translation from neuroscience to management should be treated cautiously, the underlying principle resonates with decades of organisational research: high-trust environments consistently outperform low-trust ones in measures of productivity, innovation, and employee wellbeing.

Interpersonal Versus Institutional Trust

An important distinction has emerged between interpersonal trust – the kind measured in trust games, where one individual makes themselves vulnerable to another – and institutional trust, which involves confidence in systems, organisations, and social structures. Oxytocin research has primarily illuminated the former. The biological mechanisms underlying trust in institutions remain far less understood and likely involve different neural circuits and cognitive processes.

Cross-Cultural Considerations

Most oxytocin-trust research has been conducted in Western, educated, industrialised populations – a limitation the field increasingly acknowledges. Trust norms vary dramatically across cultures. What constitutes trusting behaviour in a Zurich laboratory may carry entirely different social meanings in Tokyo, Lagos, or São Paulo. The extent to which oxytocin’s effects on trust are culturally universal versus culturally modulated remains an open and important question. Early cross-cultural work suggests that while oxytocin’s basic neurobiological functions are conserved across human populations, the social behaviours it influences are shaped substantially by cultural context.

Criticisms of the “Trust Hormone” Label

Two decades after the original trust game study, the scientific community has largely moved away from the “trust hormone” label. The reasons go deeper than failed replications.

Methodological Concerns

Gareth Leng and Mike Ludwig, in a pointed 2016 review in Biological Psychiatry titled “Intranasal Oxytocin: Myths and Delusions,” raised fundamental questions about the intranasal delivery method used in nearly all human oxytocin-trust studies. They noted that very little of the oxytocin sprayed into the nose actually reaches the brain in behaviourally relevant quantities. Instead, much of it enters the bloodstream, raising peripheral oxytocin to supraphysiological levels with potential effects on the cardiovascular system, gastrointestinal tract, and reproductive organs – effects that could indirectly influence behaviour through mechanisms entirely unrelated to central oxytocin signalling.

Furthermore, the field has grappled with publication bias. Positive, headline-grabbing results (“Oxytocin Increases Trust!”) were more likely to be published than null findings, potentially inflating the apparent strength of the effect. Many studies used small samples, tested multiple outcomes, and reported only significant results – practices that, as the broader replication crisis in psychology revealed, can produce a literature full of findings that look robust but prove fragile.

What We Actually Know

So what has survived the scrutiny? Several findings stand on relatively firm ground:

• Oxytocin plays a genuine role in social bonding. The evidence from animal research – particularly in prairie voles and rodent maternal behaviour – is extensive and well-replicated. In humans, oxytocin rises during positive social interactions including breastfeeding, physical affection, and emotional connection.
• Oxytocin modulates social attention. Neuroimaging studies consistently show that oxytocin alters amygdala reactivity to social stimuli, though the direction and magnitude of the effect depend on context, personality, and clinical status.
• The relationship between oxytocin and trust is real but conditional. Oxytocin likely facilitates trusting behaviour under certain conditions – cooperative contexts, familiar or in-group partners, safe environments – while amplifying suspicion or tribalism in others.
• Oxytocin is not a simple “trust switch.” It interacts with dopamine, serotonin, vasopressin, and other neurochemical systems in ways that are still being mapped. Its effects depend on sex, social context, individual differences in receptor density, and prior social experience.

The story of oxytocin and trust is, in many ways, a story about how science works – how a genuine and important discovery can be simplified, overhyped, partially corrected, and eventually refined into a more accurate understanding. The “trust hormone” label was never accurate, but the research it inspired has genuinely deepened our knowledge of how the brain navigates the fundamental human challenge of deciding when to cooperate with others.

Frequently Asked Questions