Technique

How to Tell If Someone Is Lying: What Actually Works (and What Doesn't)

By The MentalismPro TeamUpdated

As an Amazon Associate we earn from qualifying purchases. How we choose products.

Two silhouetted figures facing each other across a table under a single lamp

Everyone wants a quick way to tell if someone is lying, and the internet is full of lists that promise one: they look left, they touch their nose, they will not meet your eyes. Decades of research say it is not that simple. There is no gesture, no glance and no phrase that works as a lie detector on its own, and the people who claim otherwise are usually selling something.

What does exist is a smaller, more honest toolkit. Mentalists, who spend their careers watching how people react under pressure, lean on the same few ideas: learn how a person normally behaves, notice when something changes, listen to the story more than the body, and ask questions that make a made-up account harder to keep up. This guide covers each of those, says plainly where each one fails, and explains what to do once you suspect a lie.

Can You Really Tell When Someone Is Lying?

Short answer: not as well as you think. Most of us feel we have a decent instinct for it, and the research keeps showing that the instinct is only slightly better than a guess.

The 54 Percent Problem

The most cited number comes from Charles Bond and Bella DePaulo's 2006 meta-analysis in Personality and Social Psychology Review. They pooled 206 documents and 24,483 judges. When people tried to separate lies from truths in real time with no special aids or training, they were right 54 percent of the time on average. They caught 47 percent of lies and correctly accepted 61 percent of truths.

Read that second pair carefully. People are better at recognizing honesty than deception, because we start from the assumption that the person in front of us is telling the truth. The same paper found that people are more accurate when they judge what they hear than what they see, which is awkward news for anyone who relies on body language alone.

Why Confident People Still Fail

Confidence and accuracy are different things. The 2012 study by Richard Wiseman and colleagues in PLoS ONE is a small illustration: one group of volunteers was taught the popular eye-direction rules before judging video clips, and a control group was not. The two groups did not differ in accuracy or in how confident they felt. Learning a neat rule gave people a feeling of skill without any extra skill.

Add the habit of defaulting to belief, and the habit of noticing only the signs that fit a suspicion we already hold, and you get the usual pattern: a person who is sure they can spot a liar, and a result that is close to chance. The sensible response is not to give up, but to hold every impression loosely and test it.

Signs Someone Is Lying

If no single cue works, why do lists of signs of lying exist? Because some behaviors do shift a little on average when people lie. In a 2003 review in Psychological Bulletin, Bella DePaulo and colleagues combined 1,338 estimates of 158 cues to deception. Liars tended to be less forthcoming, told less compelling tales, made a more negative impression and were more tense, and their stories had fewer ordinary imperfections and unusual details. The same review also found that many behaviors showed no discernible link, or only a weak link, to deceit. Cues were stronger when people were highly motivated to succeed and when the lie concerned a transgression.

Notice what kind of signs those are. They are tendencies across many people, not tests you can run on one person. Each sign below comes with its limit.

Changes From Baseline

This is the idea mentalists and interviewers return to most. Before judging anything, learn how the person acts when they are relaxed and honest: how fast they talk, how much they gesture, how often they look away, how they sound when they joke. A sudden change when you reach a sensitive topic, such as a person who is usually animated going flat, or a person who is usually quiet becoming talkative, is worth noticing.

The limit is that stress changes behavior whether or not someone is lying. An honest person who fears being disbelieved can look every bit as nervous as a liar, and some liars feel nothing at all. A baseline change tells you the topic matters to the person. It does not tell you why.

To use it well, compare like with like. Judge how someone acts on a hard topic against how they act on another hard topic, not against how they act when you chat about the weekend. The principle is the same one behind the guide onhow to read people: a single behavior means little, a pattern of change means more.

Story Details and Timeline

The DePaulo review points to the story rather than the body: liars tend to be less forthcoming and to tell less compelling tales. In practice, listen for the texture of the account. Real memories usually contain sensory detail, side issues, small complications and things that did not go to plan. Invented stories are often tidy, short on specifics, and built around the point the person wants to land.

Check the timeline too. Does the order of events hold together? Can they say what happened just before and just after the key moment? A fabricated account is easy to tell at the headline level and harder to extend in every direction.

The limit: some honest people simply tell short, plain stories, and some real memories are patchy, especially after a stressful event or a long time. A thin story is a reason to ask for more, not a finding of guilt. Equally, a very polished, rehearsed story can be a warning, but it can also come from someone who has told the same story many times.

Voice and Pauses

Bond and DePaulo found judges did better with what they heard than with what they saw, so voice is worth attention. Listen for a change in pace, a rise or drop in pitch, longer gaps before answering, or more false starts than usual compared with the person's baseline.

The limit is large: people also pause to remember, to choose their words carefully or to avoid hurting you. Anxiety, a bad phone line and a second language all change voices. Use a pause as a cue to ask a follow-up, such as "Tell me a bit more about that," and then see whether the detail arrives.

Over-Control and Rehearsed Behavior

Many people think liars fidget, so liars sometimes try to stay very still and keep their face neutral. That effort can show up as stiffness, as fewer natural gestures than usual, or as a performance that looks a little too smooth. A person who is unusually composed for a topic that would normally rattle them is, again, a change from baseline.

The limit is the mirror image of the first sign. Plenty of honest people are calm, practiced speakers, and plenty of liars are relaxed. Over-control is a reason to look at other things, not to conclude anything. If you want the broader picture of why gestures are such unreliable evidence, see the section on myths in the guide toreading people.

How to Tell If Someone Is Lying by Their Eyes

This is one of the most searched versions of the question, and the one with the most myths attached.

The Eye Direction Myth

You will often read that eyes moving up and to one side mean a person is making something up, while eyes moving to the other side mean they are remembering. That idea comes from neuro-linguistic programming, and it has been tested. In the 2012 PLoS ONE paper, Wiseman and colleagues ran three studies. In the first, with 32 right-handed adults who either lied or told the truth in filmed interviews, lying and truth-telling did not produce different patterns of upward gazes. In the second, 50 people judged those clips, and teaching one group the eye-direction claims did not make them any more accurate. In the third, the researchers coded 52 videos of real public pleas for missing relatives, where the lies were established by strong evidence, and again found no significant differences.

The authors concluded that the three studies provided no evidence for the claim. So the answer to the popular question about eyes left or right is plain: neither direction tells you someone is lying.

Eye Contact and Blinking

Two other beliefs sit beside it. The first is that liars avoid eye contact. Some do, many do not, and many honest people look away constantly because looking away helps them think, or because they are shy, or because of their culture. Some liars deliberately hold your gaze for longer than normal precisely because they know people look for the opposite.

The second is that blinking, pupil changes or a quick glance down expose a lie. These are tiny, fast movements that observers cannot read reliably in real conversation, and the research above gives no reason to trust them. The same goes for the fleeting facial flashes covered in our guide to micro expressions. The practical rule for eyes is the same as for everything else: compare against the person's baseline, treat any change as a prompt to ask a question, and do not decide anything from the eyes alone.

Questions That Expose a Lie

If watching is unreliable, the better lever is what you ask. The research tradition on this is associated with psychologist Aldert Vrij and colleagues, who study how to make lying harder rather than how to spot a twitch. Telling a convincing lie takes mental effort, and the idea is to add more effort at exactly the point the liar is least prepared for.

Unexpected Questions

A prepared liar has an answer ready for the obvious questions: where were you, who were you with, what did you do. They are less ready for the sideways ones. Ask about something they could not have rehearsed: "What was the first thing you noticed when you walked in?" or "What were you thinking about on the way there?" A person recalling a real event answers from memory. A person building one has to invent on the spot while keeping the rest consistent.

The limit: an unexpected question can fluster an honest person too, and memory for small details is imperfect. Judge whether the answer is plausible and consistent with the rest of the account, not whether it arrived instantly.

Tell It Backwards

Reverse-order recall is one of the best-known techniques. In a 2008 study by Vrij and colleagues in Law and Human Behavior, 80 mock suspects told true or false stories, some of them in reverse order. Accounts told in reverse contained many more cues to deceit, and when 55 police officers judged the recordings, reverse-order recall improved their accuracy without making them more suspicious across the board.

The logic is simple. A true event is stored as a memory you can walk through in either direction. A false story is usually built in forward order, so running it backwards is hard. Ask, "Can you start from the end and walk me back through it?" Then listen for gaps and contradictions. As with all the work in this area, those were laboratory studies with mock suspects, so real life will be messier.

Ask for Verifiable Detail

A lie that cannot be checked is cheap. A lie that can be checked is expensive. Ask for details you could confirm in principle: who else was there, what time something happened, which route they took, what was on the receipt, what the message said. You do not need to confirm all of it. Often the willingness to be specific, and the manner in which they do it, tells you more than the detail itself.

Keep your tone curious rather than accusing. People who feel interrogated clam up whether or not they are honest, and you learn less.

The "3 Questions" and "10 Second Trick" Answers

People often search for a 10 second trick to spot a liar, or for how to spot a liar in 3 questions. There is no 10 second trick that works, and any article offering one is overselling. What you can do in a short conversation is combine the ideas above into three questions:

  1. An open invitation: "Walk me through what happened." Listen for texture and timeline.
  2. An unexpected question: "What was the first thing you noticed?" See how the answer compares with the first.
  3. A checkable detail or a reverse pass: "Who else could tell me that?" or "Start from the end and go back."

Three questions will not prove anything. They will give you better material than any amount of staring.

Phrases Liars Use

Searchers often ask what phrases liars say. The honest answer is that no phrase is a confession, but a few patterns are worth noticing, as long as you compare them with how the person normally speaks.

  • Distancing language: avoiding "I" and "my," or referring to "that woman" instead of a name they normally use. It can reflect a wish to detach from the story.
  • Oaths and assurances: "I swear," "honestly," "to be perfectly frank." Honest people say these too, and a habit of using them is part of someone's baseline. A sudden pile of them on one subject is more interesting.
  • Repeating your question: "Did I take the money?" Repeating a question buys time. It is also what careful people do when they want to be sure they understood.
  • Answering a different question: a smooth reply that sounds relevant but does not address what you asked. It is worth noticing when it happens once, and more worth noticing when it keeps happening on the same topic.
  • Too few details, or too many: a vague account that avoids specifics, or an over-explained one full of unrequested justification.

Treat these as flags that tell you where to ask the next question. They are weaker evidence than the story itself, and each has plenty of honest explanations.

How to Tell If Someone Is Lying Over Text

Texting removes the body entirely, which, given how unreliable body cues are, is less of a loss than it sounds. What you lose is voice, which the 2006 meta-analysis suggests is among the more useful channels. What you gain is time and a written record.

Useful things to look at:

  • Baseline first: how quickly does this person usually reply, how long are their messages, do they use punctuation, emoji, nicknames? A sudden shift on one topic is the thing to notice, not any single habit.
  • Reply speed: a very fast reply to a hard question can mean a prepared answer, and a slow one can mean someone is composing. Or it can mean they were in a meeting. Speed alone proves nothing.
  • Vagueness: messages that dodge specifics, answer a different question, or change the subject.
  • Over-explanation: long unprompted justifications, especially when you only asked a simple question.
  • Consistency: because text is written down, you can compare what they say today with what they said last week. That is a stronger check than any tone-of-voice guess.

Over text, the best tools are the same as in person: ask for detail that can be checked, and ask something they have not prepared for. A short call or a face-to-face talk for sensitive topics gives you voice and pauses back, and it makes a spontaneous story harder to maintain than a drafted message.

How Mentalists Use This on Stage

Mentalism is an art of entertainment, and the skills overlap with lie detection only partly. A performer who works close up pays attention to small reactions, to hesitation and to what a volunteer chooses when they believe no one is guiding them. That is the same attention you would use in a conversation.

The classic "liar and truth-teller" routine is a good example of how the theater works. A volunteer hides something or makes a choice, and the performer announces which statements were true and which were false. It is staged to look like a feat of reading faces. Audiences tend to assume the performer is seeing tiny signs of deception, and a careful performer will often tell you honestly that no one can do that reliably. The craft is in the structure of the routine and in the performer's control of the situation, not in a superhuman lie detector, and entertainers who claim real powers are being showmen, not scientists.

The takeaway for everyday life is humbling: even people who spend their lives watching others are not reading minds. If you want to understand the techniques behind the theater, start with cold reading, which explains how performers appear to know more about a stranger than they do, and the wider overview of mentalism techniques. For a lighter, practical way to practice noticing reactions, the collection of mind tricks is a good place to try things on friends who know they are part of a game.

What to Do When You Know They Are Lying

Suppose the evidence has stacked up. The question that people search for next is what to do when you know someone is lying but they will not admit it. The honest answer starts with how sure you actually are.

  1. Separate suspicion from knowledge. If your certainty comes from signs, remember the 54 percent figure. If it comes from a fact you can document, such as a contradicting message or a receipt, you are on firmer ground.
  2. Decide what you actually want. An admission, a changed behavior, safety, an end to the relationship, or simply the truth for your own peace of mind? The answer changes the approach.
  3. Pick a private, calm moment. Public confrontation and heat make people double down. State what you observed rather than what you think they are: "You told me you were there at six, and the message says seven. Help me understand."
  4. Leave them an exit. People who can save face are more likely to correct the record. Someone who feels cornered will lie again.
  5. Accept that an admission may not come. You may never get one. You can still act on the facts, set a boundary, or take a decision without it.
  6. Look after yourself. If the lie is about something serious, such as money, safety or health, or you have a legal or workplace issue, document it and ask a professional, a manager or a lawyer for advice instead of running your own interrogation.

A Book on Reading Behavior

  • Best for reading people

    What Every Body Is Saying

    Joe NavarroBeginner

    A former FBI agent's guide to nonverbal behavior. Read it for the emphasis on baselines and clusters of behavior, and keep in mind that the lie detection research above is more cautious than any book.

    Check price on Amazon

There is no 10 second trick that works. Research found that people judge lies and truths correctly only about 54 percent of the time. The nearest thing to a shortcut is to notice a change from how the person normally behaves, then ask one unexpected question or ask them to tell a short part of the story backwards.
No phrase proves a lie. Patterns worth noticing include distancing language, a sudden pile of oaths such as "I swear," repeating your question to buy time, answering a different question, and vague or over-explained accounts. Compare them with how the person normally talks before reading anything into them.
Start with an open invitation, such as "Walk me through what happened." Then ask an unexpected question they could not have rehearsed, like "What was the first thing you noticed?" Finish with a checkable detail or a request to go back through the event in reverse order. None of these proves a lie, but each makes a made-up story harder to keep consistent.
First check whether you know or only suspect. Then decide what you want from the conversation, choose a calm private moment, describe what you observed instead of attacking, and leave them a way to save face. If no admission comes, you can still act on the facts and set your own boundary.
No. The idea that looking to one side signals a lie comes from neuro-linguistic programming. A 2012 study by Wiseman and colleagues in PLoS ONE tested it in three studies, including real public appeals for missing relatives, and found no evidence for it.
Not with certainty. A polygraph records body responses such as heart rate and sweating, which also change with fear and stress, so the result is an indication that needs interpretation rather than proof of a lie. Treat any claim of a perfect lie detector with suspicion.

Free newsletter

Mentalism, made modern. In your inbox.

Occasional emails from the MentalismPro team: a technique to practice, a routine idea, and the new guides and books worth your time. No spam, unsubscribe any time.