Case File 075 · Investigation and Psychology

What Is Behavioral Leakage?

The origin of the idea that hidden feelings leak through nonverbal behaviour, what the research found, and why leakage is not proof of deception.

By Murder Book Editorial TeamPublished 16 September 2026Updated 16 September 2026

Behavioural leakage (spelled "behavioral leakage" in American English) is the idea that feelings a person is trying to hide can slip out through their nonverbal behaviour, such as a brief facial expression, a gesture or a change in posture. The concept comes from psychologists Paul Ekman and Wallace Friesen, whose 1969 paper "Nonverbal leakage and clues to deception" offered the first theoretical account of how nonverbal behaviour relates to deception. Research since then has found that concealed emotions do sometimes show briefly. It has also found that behavioural cues to lying are weak and unreliable, that honest people show many of the same signs, and that observers judge lies and truths only slightly better than chance. Leakage can suggest that someone is feeling something. It cannot show why, and it is not evidence that a person is lying.

Key takeaways

Ekman and Friesen proposed in 1969 that concealed feelings can "leak" through nonverbal behaviour, and that people control some parts of the body more carefully than others.
A 2008 laboratory study found brief, inconsistent emotional expressions when participants faked or masked emotions, but full microexpressions appeared in only 2% of expressions.
Meta-analyses find that many behavioural cues to deception are weak or absent, and people average 54% accuracy when judging lies and truths.
Anxiety does not separate liars from truth tellers, because honest people can be anxious about not being believed.
A U.S. government review found no scientific consensus that behaviour detection can reliably be used for counterterrorism screening.

Where does the idea of behavioural leakage come from?

Paul Ekman and Wallace Friesen published "Nonverbal leakage and clues to deception" in the journal Psychiatry in 1969. According to a 2024 introduction to a special issue of the Journal of Nonverbal Behavior by Sally Farley, it was the first theoretical formulation of nonverbal behaviour and deception, and it sparked decades of research.

Farley notes that the paper drew on Sigmund Freud's view that a person who stays silent may still betray themselves through their movements. Ekman and Friesen developed this into what is often called the leakage hierarchy. It held that the behaviours we are practised at controlling, such as the face, reveal less about possible deception than behaviours we pay less attention to, such as the legs and feet.

A 2024 study by Li and colleagues summarises the core idea as the nonverbal leakage hypothesis: when people attempt to deceive, they may show nonverbal behaviour linked to emotions such as anxiety or fear. Farley notes that the theory has intuitive appeal and has been profoundly influential in popular media and television, but that it has come under attack in recent years.

What does leakage look like in research?

One of the best-known tests of emotional leakage in the face is a 2008 study by Porter and ten Brinke, published in Psychological Science.

Participants viewed disgusting, sad, frightening, happy and neutral images. For each one, they responded with either a genuine expression or a deceptive one: simulating an emotion they did not feel, neutralising their expression, or masking what they felt. The researchers analysed 697 expressions frame by frame, a total of 104,550 frames.

The study reported several findings:

Masked emotions were associated with more inconsistent expressions and a higher blink rate than genuine emotions.
Negative emotions were harder to fake convincingly than happiness.
Inconsistent emotional leakage appeared at least once in every participant and lasted longer than the standard definition of a microexpression.
True microexpressions, lasting 1/25 to 1/5 of a second, were shown by 21.95% of participants in just 2% of all expressions, and only in the upper or lower half of the face.
Untrained observers performed only slightly above chance at detecting the deception.

So leakage in this sense is real, but it is brief, patchy and easy to miss. The study also took place in a laboratory, using emotional images rather than lies about real events.

Why is leakage not proof that someone is lying?

This is the most important limit of the concept, and several lines of research point the same way.

Behavioural cues to deception are weak. A 2003 meta-analysis by DePaulo and colleagues combined 1,338 estimates of 158 possible cues. Liars were, on some measures, less forthcoming and more tense. But many behaviours showed no discernible links, or only weak links, to deceit.

People are poor lie detectors. Bond and DePaulo's 2006 meta-analysis of 24,483 judges found an average of 54% correct lie-truth judgments. A later lens-model meta-analysis by Hartwig and Bond (2011) concluded that the main problem is not that people rely on the wrong cues, but that the behavioural cues themselves are weak.

Emotion is not the same as deception. Farley's 2024 review notes that a key problem for leakage theory is that anxiety does not discriminate truth tellers from liars. A liar may be anxious about being caught. A truth teller may be anxious about being wrongly labelled a liar. A visible emotion does not tell you which of these is happening, or whether the cause is something else entirely, such as grief, embarrassment or fear of the situation.

Some reported effects may be exaggerated. Luke (2019) used simulations to show that many estimated effect sizes for deception cues may be greatly inflated by publication bias, small numbers of estimates and low statistical power. He warned against relying on potentially illusory cues.

Leakage does not always appear. In a series of experiments on equivocal communication, Bavelas, Black, Chovil and Mullett (1990) reported that nonverbal leakage did not occur in either equivocal or false messages.

A 2019 Annual Review of Psychology article by Vrij, Hartwig and Granhag summarised the field this way: the nonverbal cues to deceit discovered so far are faint and unreliable, and people are mediocre lie catchers when they pay attention to behaviour. The authors also noted that misbeliefs about nonverbal behaviour and deception appear very hard to debunk. For a broader look, see Why There Is No Single Reliable Sign of Deception.

What happens when behaviour reading is used in real settings?

The limits matter most when behaviour is used to make decisions about people.

In 2010, the U.S. Government Accountability Office (GAO) reviewed the Transportation Security Administration's SPOT programme, which screened airline passengers by observing their behaviour. The GAO reported that "a scientific consensus does not exist" on whether behaviour detection principles can be reliably used for counterterrorism purposes, citing the National Research Council. It also noted that the programme had been deployed nationwide before its scientific validation was established.

Researchers have also warned about unsupported claims reaching the justice system. Jupe and Denault (2019) argued that some alleged experts have given police officers, lawyers and judges information about nonverbal communication that has not been empirically tested or supported by scientific theory.

How does leakage compare with other approaches?

Farley's 2024 review describes cognitive theories as the most promising alternative to the leakage hierarchy. Instead of looking for escaped emotion, these approaches start from the idea that lying can be more mentally demanding than telling the truth. Farley reports a meta-analysis of 14 studies in which approaches that increased cognitive demands raised lie detection accuracy from 47% to 67%. She also notes that these approaches mainly produce differences in verbal cues, not nonverbal ones.

Even here, accuracy remains far from certain. What Is Cognitive Load in an Interrogation? explains the idea in more detail.

Popular belief	What research suggests
Hidden feelings always show on the body	Brief inconsistent expressions can occur, but microexpressions were rare in a key laboratory study
Nervous behaviour means lying	Anxiety does not separate liars from truth tellers
People can spot leakage reliably by watching	Average accuracy is about 54%, and cues themselves are weak
Strong deception cues are well established	Many reported effect sizes may be inflated
Behaviour screening is scientifically validated	A GAO review found no scientific consensus for its counterterrorism use

Can you tell if someone is lying from their behaviour?

Not reliably. Behaviour can tell you that something may be worth asking about. It cannot tell you what that something is. The practical lesson from leakage research is to treat an unexpected expression as a question, not an answer, and to look for independent evidence. Can You Actually Tell When Someone Is Lying? looks at the evidence on lie detection as a whole.

How does Murder Book use emotional leakage?

Murder Book: The Art of Interrogation is an upcoming narrative interrogation game from Gungrounds, an independent studio in Zagreb, Croatia. Its press kit names Emotional Leakage among the interrogation behaviours players examine, alongside Fabrication, Avoidance and Narrative Gaps. Development screenshots show an in-game casebook with entries such as Stonewalling, Stress Response, Acting Emotion and Emotional Fading. The Acting Emotion entry links failing to sustain a performed emotion to cognitive load theory.

The research above explains why those entries work best as prompts rather than verdicts. A flash of emotion might mean concealment, but it might also mean fear, grief or the stress of being questioned. In Murder Book, behaviour is a reason to look closer, not proof. Cases are designed for focused 5–10-minute sessions.

Murder Book is in development for PC (Steam) and mobile, with release currently planned for the end of Q2 2027. If you enjoy the deception side of detective games, see Best Games About Catching Lies.