Case File 211 · Investigation Psychology

Can Body Language Tell You If Someone Is Lying?

By Murder Book Editorial TeamPublished by Gungrounds, the studio making Murder BookPublished 3 October 2026Updated 3 October 2026

No. Body language cannot tell you reliably whether someone is lying. The popular signs — avoiding eye contact, fidgeting, crossed arms, touching the face — are not dependable indicators of deception. Decades of research find that behavioural cues to lying are weak, that honest people show the same behaviours when they are nervous or uncomfortable, and that observers who watch for these cues judge truth from lies at only about 54% accuracy, barely above a coin toss. There is no bodily "tell" that works. The strongest finding in the field is a negative one: the confident belief that liars give themselves away through their body is itself the most robust and most mistaken part of the picture. Body language can show that a person is uneasy. It cannot show why.

Key takeaways

There is no reliable body-language sign of lying; the familiar cues do not separate liars from truthful people.
People average about 54% accuracy at detecting lies, close to chance, whether or not they watch body language.
Behavioural cues to deception are weak overall, and honest people often show the "suspicious" behaviours out of nervousness.
The belief that liars avoid eye contact is the most common deception stereotype in the world, and it is wrong.
Reading body language as proof risks treating anxiety, discomfort or cultural difference as guilt.
Nonverbal behaviour is, at most, a reason to ask another question, never evidence that someone is lying.

Why do people believe body language reveals lies?

The belief is nearly universal. In a landmark cross-cultural study, the Global Deception Research Team asked people in 75 countries how to tell when someone is lying. The single most common answer, everywhere, was that liars avert their gaze. Fidgeting, self-touching and shifting posture also featured heavily.

The stereotype is intuitive and culturally reinforced, taught through films, folklore and some training courses. But its popularity is not evidence that it works. The same study and the wider literature show that these shared beliefs are largely inaccurate: the behaviours people are most confident about, such as gaze aversion, have little or no real association with lying. A belief can be universal and still be false.

What does the research actually find?

Three findings, repeated across large studies, undercut the idea that bodies betray lies.

First, the cues themselves are weak. DePaulo and colleagues' 2003 meta-analysis pooled 1,338 estimates covering 158 possible cues to deception. Most showed no meaningful link to lying, and the links that did appear were faint. There was no behavioural signature of deceit.

Second, people are poor lie detectors. Bond and DePaulo's 2006 meta-analysis of nearly 25,000 judges found an average accuracy of 54% — only marginally better than guessing. Watching for nonverbal cues did not rescue performance.

Third, the problem is not simply that people watch the wrong cues. Hartwig and Bond's 2011 lens-model meta-analysis concluded that human lie detectors fail largely because the available behavioural cues are weak to begin with, not because observers are merely looking at the wrong things. Even an ideal observer would not have much to work with. A 2019 review in the Annual Review of Psychology by Vrij, Hartwig and Granhag summarised the field bluntly: the nonverbal cues to deceit discovered so far are faint and unreliable.

Why do honest people look nervous?

Because the behaviours blamed on lying are mostly signs of stress, discomfort or concentration, and those have many causes. An innocent person who is being accused, who fears not being believed, or who simply finds authority intimidating can fidget, avoid eye contact and tense up — exactly the pattern a naive observer reads as guilt.

This is the core reason body language fails as a lie test: the signal is ambiguous at its source. Anxiety does not distinguish a liar from a truthful person who is anxious about being disbelieved. Cultural norms add another layer; steady eye contact is expected in some cultures and considered rude or confrontational in others, so "avoiding eye contact" can mean nothing more than politeness. Reading nervousness as deception turns ordinary human discomfort into false suspicion. The concept of behavioural leakage — the idea that hidden feelings slip out nonverbally — captures what is real here, and also why it falls short as proof.

Are microexpressions or specific "tells" any better?

They are far weaker than their reputation. Brief, involuntary expressions do occur, but in controlled studies full microexpressions are rare and appear in only a small fraction of expressions, often in people who are not even lying. No single gesture — a touched nose, a glance to one side, a swallowed word — has been shown to reliably indicate deception. Claims that a particular tell exposes liars tend to outrun the evidence, and some have reached courtrooms and training programmes without scientific support, which worries researchers precisely because the stakes are high. For the broader case, see why there is no single reliable sign of deception.

So is body language useless in an interview?

Not useless, but easily misused. Nonverbal behaviour can flag a moment worth exploring: a question that suddenly produces discomfort might be worth returning to. The error is jumping from "that reaction was interesting" to "therefore they are lying." The responsible use is diagnostic humility — treat a behaviour as a prompt for another question or for seeking independent evidence, not as an answer.

Approaches that work better tend to shift attention away from the body and toward the content of what is said, for example by increasing the mental difficulty of lying or by checking accounts against verifiable facts. Even these raise accuracy only modestly. If you want the wider view of what is and is not possible, can you actually tell when someone is lying? weighs the evidence, and what makes a statement unreliable looks past behaviour to the account itself.

How does a game handle body language without faking certainty?

Murder Book: The Art of Interrogation is built on the same restraint the research demands. It is an upcoming narrative interrogation game from Gungrounds, an independent studio in Zagreb, Croatia, inspired by real police footage and psychological interviews. Its in-development casebook lists behaviours such as Stress Response, Submissive Gesture and Acting Emotion, and its design states plainly that recognising a behaviour is never treated as proof that a suspect is lying or guilty.

In the game, noticing how a suspect reacts can open new questions and branches, but it does not hand you a verdict. A flicker of stress might mean concealment, or fear, or the pressure of the room — and the case stays open to several explanations until the evidence, not the body language, settles it. 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.