How to Read People: 7 Body Language Cues Decoded

Learn how to read people with 7 body language cues, the baseline method that makes them work, and where the cues stop being reliable.

Reading people is the ability to observe and interpret nonverbal cues (facial expressions, posture, gestures, and vocal tone) to understand what someone is thinking or feeling beyond their words. The method is learnable, and it’s one of the most useful things you can practice at work.

As Vanessa Van Edwards puts it in Cues: “The biggest mistake people make is treating body language like a dictionary—one gesture, one meaning. Real people-reading requires observing clusters of signals and comparing them to a person’s baseline behavior.”

And people are not good at the dictionary version: trying to spot whether a stranger is lying, they get it right about 54% of the time¹, barely above a coin flip.

The Golden Rule: Baseline First, Then Read

A baseline is how someone normally behaves when they’re relaxed and have nothing to hide. Some people always cross their arms. Some people naturally avoid eye contact. Some people fidget constantly. None of that means anything until you know what’s different for that specific person.

The first few minutes of any interaction are for watching. How does this person sit? Where do their hands rest? How much eye contact do they make, and what does their voice sound like when they’re talking about something neutral: the weather, or how their weekend went?

That’s your baseline. Everything after that is about spotting deviations from it.

It’s also why “crossed arms means closed off” survives as long as it does. Crossed arms are a resting position for a lot of people. On someone who never crosses them, the same gesture is worth a second look.

Action Step: The next time you’re in a meeting or conversation, spend the first two or three minutes just observing. Ask yourself: “What does relaxed look like for this person?” Make a mental note. Then watch for shifts.

No single gesture means anything on its own. You need to know what’s normal for that specific person before you can spot what’s different.

The Micropositive and Micronegative Framework

Once you have a baseline, you need a system for sorting what you see. Every body language cue falls into one of two buckets:

• A micropositive shows interest, curiosity, or engagement.
• A micronegative shows nervousness, disinterest, or discomfort.

In any interaction, you want to see more micropositives than micronegatives. Never rely on a single cue. Look for clusters: multiple signals happening at the same time across different channels (face, body, voice, words).

Someone breaking eye contact has told you nothing on its own. Add a drop in volume and shoulders angling toward the door, and now you have something worth acting on.

Across 66 behaviors in 153 samples, researchers found that people rarely watch the wrong signals. The signals themselves are just weak². Which is why several signals together beat any one of them on its own. Combinations contain real information: fed every cue and measured precisely, statistical models identified deception about 68% of the time³.

Here are the cues worth watching for.

Cue #1: Spotting Shame and Embarrassment

Show people a particular configuration and they call it shame far more often than chance: gaze directed downward, and the head dropping with it⁴.

This is a micronegative. You see it in meetings when someone realizes they made an error, or in conversations when a sensitive topic comes up.

Embarrassment looks slightly different: it often includes a non-Duchenne smile (a smile that doesn’t reach the eyes), a lip press, gaze down, and sometimes touching the face or forehead. That forehead touch is a self-soothing gesture, a starter movement for wanting to hide or block out what’s happening. If someone is really embarrassed, the forehead touch can turn into a full eye block, where they go from touching their forehead to covering their eyes entirely.

Observers rarely mix up these two. In the same study, observers shown the embarrassment display called it shame only 7% of the time, and shown the shame display called it embarrassment 3.4% of the time⁴. Two configurations that look similar on paper are separable in practice, which is why it pays to know which one you’re looking at. Embarrassment usually wants rescuing with a laugh. Shame usually wants the subject changed.

Action Step: Watch for the gaze dropping and the head going down with it. If a hand comes to the face and the smile goes tight, you’re looking at embarrassment instead. Either way, it’s a moment to ease up or change the subject.

Want the full catalogue? Cues works through 96 of the signals people send each other, from stance and facial expression to word choice and vocal tone, and what to do with each one.

Cue #2: Blocking Behavior

Blocking is when someone covers part of their body, creating a barrier between themselves and another person. This might look like crossing arms, crossing legs, or holding something like a laptop, a notepad, or a pillow in front of their torso. Most people do it without noticing, and the usual explanation is self-protection.

It’s a micronegative, and it’s the one readers ask about most.

The key is context. Someone crossing their arms while frowning and leaning away? Likely disengaged. Someone crossing their arms while smiling and maintaining eye contact? Probably just comfortable. And someone crossing their arms in a cold conference room? They might just be chilly.

It looks like this in practice. You mention the timeline. Their arms come up and lock across the chest. The feet, which were pointed at you, shift toward the door. The answer comes later than the last three answers did, and it’s shorter. That’s four signals across three channels, all pointing the same way, inside about two seconds. Any one of them alone would be nothing.

Crossed arms don’t automatically mean someone is closed off. Look for tight gripping of the arms versus relaxed crossing before drawing conclusions.

I found the most cringeworthy example of this in an old Blind Dating episode that went terribly wrong. I have to show you their first impression. Notice any blocking behavior in this video?

Did you see how she crossed her arms and legs right from the start? Combined with her facial expression and body orientation, that cluster of signals painted a clear picture. It didn’t end well.

Action Step: When you notice blocking, check for clusters before drawing conclusions. Ask yourself: Are their arms tightly gripped (tension) or loosely crossed (comfort)? What’s their face doing? Are they leaning toward me or away? The arms alone tell you almost nothing.

Cue #3: The Head Tilt

Do you hear that? It’s a natural human behavior to tilt our head and expose our ear when we want to hear something better. This one is a micropositive.

If someone tilts their head while you’re speaking, it’s usually a sign of engagement. The nod that usually comes with it belongs to a family of small movements whose job is keeping a conversation running, alongside the slight lean forward and the brow raise⁵.

You know who is great at this? The ultimate great listener, Oprah Winfrey. I think one of the ways Oprah gets people to open up to her is with her amazing nonverbal. Let’s watch this clip and see how she listens with her entire body:

Did you catch the slight head tilts paired with subtle nods? The head tilt plus nod is one of the easiest micropositives to spot. It communicates: “I hear you, keep going.”

Action Step: If someone tilts and nods while you speak, take it as encouragement to keep going, and look for whether their questions and follow-ups agree with it. Want to show someone you’re listening? Give them the head tilt and nod. It works in meetings, one-on-ones, and even video calls.

Cue #4: The Mouth Cover

Have you ever seen a little kid tell a lie?

Watch one long enough and you’ll see a lie followed by a hand going straight to the mouth, as if the brain were saying, “No, don’t say it!”

Adults do a subtler version of this. A mouth cover tends to go with stress, and stress is not the same as a lie. There’s even a name for the mistake of assuming otherwise, coined by the researcher whose work mapped these expressions: the Othello Error, after Shakespeare’s Othello, who concluded Desdemona was lying because she looked distressed. She was distressed because she was being falsely accused.

Practitioners usually file the mouth cover as a pacifying behavior: a way the body manages discomfort, surprise, or mental effort. It can mean someone is holding something back, but it can also mean they’re embarrassed, concentrating, or simply processing something unexpected.

Lady Gaga does it. Howie Mandel does it. Even Sharon Osbourne does it.

Action Step: When you spot a sudden mouth cover, don’t jump to “they’re lying.” Instead, note the context. What was just said? Did the topic shift? Look for other cues in the cluster, like voice pitch changes, a shift in posture, or distancing language, before drawing any conclusions.

Cue #5: Hand Gestures

Hands tell us so much about what a person is thinking. In fact, I could show you a clip on silent, and you would know what the person is saying without hearing a single word: What you talking about? Loooooove you! Look at my face!

We instinctively watch someone’s hands as they speak. A widely used classification sorts nonverbal behavior into five categories: emblems, illustrators, affect displays, regulators, and adaptors⁵. Three of them cover most of what the hands are doing in an ordinary conversation:

1. Emblems: Gestures with specific, agreed-upon meanings (thumbs up, peace sign). These are culture-specific, so be careful. The ring gesture, thumb and index finger touching, means “A-OK” in the US and “money” in other places, and a gesture that is friendly in one country can mean something very different in the next.
2. Illustrators: Hand movements that accompany speech to emphasize or clarify a point. They tend to rise and fall with enthusiasm. A sudden decrease often means the person’s attention has gone somewhere else mid-sentence. That shift is the thing to notice.
3. Adaptors: Self-soothing touches like hair twirling, pen clicking, or face touching. They’re the hardest of the three to read, because the same movement shows up in comfort and discomfort alike.

More hand gestures are a great micropositive because they help keep people engaged and make your message more memorable.

Action Step: Use more hand gestures when you speak. They’re a natural engagement booster. And watch for sudden changes in someone else’s gesture patterns. If a normally expressive talker suddenly goes still-handed, something shifted internally. Learn the full set of hand gestures and what they mean:

Cue #6: The Eyebrow Raise

Eye contact gets most of the attention, and the eyebrows get almost none. The eyebrow raise is a great micropositive.

When someone is interested or curious, their eyebrows often go up. It’s almost as if the brain wants to clear the way so the eyes can take in more information. You see it when someone hears surprising news, encounters an attractive person, or becomes genuinely curious about what you’re saying.

The eyebrow flash, a quick raise most people never consciously notice, turns up as a greeting across widely separated cultures. Filming 255 of them in unstaged conversation across three cultures, researchers found the same movement pattern everywhere, most often paired with a smile, working as a kind of social marker that stamps meaning onto whatever else is happening⁶. When someone flashes their eyebrows at you from across a room, they’re saying: “I see you, and I’m open to interaction.”

One footnote on where the eyebrow belongs in the scheme above: brow raises were filed with the regulators, the traffic signals of conversation, alongside the head nod and the small lean forward.

Action Step: Raise your eyebrows slightly when someone says something interesting. It shows genuine engagement. And don’t miss an eyebrow raise directed at you. It usually means the other person is paying attention and wants to connect.

Cue #7: What the Face Actually Tells You

The face is the cue people most want to decode, and the one most often oversold. Depending on which expression you see, facial expressions can be both micropositives and micronegatives.

Seven expression categories are worth learning as vocabulary. These descriptions are the working shorthand nonverbal trainers use, not findings in their own right:

1. Anger: Lowered brows drawn together, hard stare, tensed lower lids, lips pressed firmly or squared open
2. Contempt: One-sided mouth raise or tightening
3. Disgust: Nose wrinkled, upper lip raised
4. Fear: Brows raised and drawn together, eyes wide, mouth slightly open
5. Happiness: Cheeks raised, crow’s feet around eyes (called a “Duchenne smile” when genuine)
6. Sadness: Inner brow corners raised, lip corners pulled down
7. Surprise: Brows raised high, eyes wide, jaw dropped

Researchers put that idea to a test: when someone actually feels one of these emotions, how often does the matching face turn up? Where the whole expression had to be present and precisely coded, about 13% of the time, and the studies clearing that bar covered only disgust, happiness, and surprise. Allow partial expressions and looser coding across all six and it reaches roughly 23%⁷.

Their conclusion is blunt: the classic basic emotions do not dependably arrive with the faces we expect from them. Basic-emotion researchers challenged that analysis in print, and the authors’ published reply restated it from the other side. Where the complete expression was required, it was missing in 87% of cases. Where the coding was looser, it was still missing in 77%⁸.

A separate review of the whole field reached the same place. Facial movements are poor diagnostics for specific emotions, and the same scowl turns up in situations that have nothing to do with anger⁹. Read the list above against the situation in front of you.

Contempt still deserves attention. In the working shorthand it’s the asymmetric one, showing on one side of the face as a subtle smirk or half-smile, and it turns up constantly in relationship research, usually attached to a headline figure about predicting divorce. Those figures are worth treating carefully. Researchers took an equation that sorted couples with about 90% accuracy on the data it was built from, then ran it on a fresh sample. Accuracy fell to 69%. When it flagged a couple as heading for divorce, it was right about 29% of the time¹⁰. Contempt is worth watching for, and worth watching twice.

Contempt tends to show on one side of the face, as a subtle smirk or half-smile. Notice it. Then keep watching.

Pro Tip: The difference between a real smile and a fake one is in the eyes. A genuine smile creases the skin around the eyes, and most people find that part hard to produce on demand. If someone smiles at you and their eyes stay flat, the smile is merely social.

For the face specifically, microexpressions goes through the seven in far more detail than there’s room for here.

Bonus Cue: Listen to the Voice

Most people-reading advice focuses on what you can see. But across five experiments, listeners read emotions slightly more accurately from voice alone than from seeing the face and hearing the voice together¹¹.

Why? Faces are easier to fake. We spend our whole lives learning to manage our facial expressions, forcing smiles in meetings and putting on a “brave face.” The voice is much harder to control. Subtle tremors, pitch shifts, and changes in breathing are linked to involuntary nervous system responses.

Closing your eyes on a phone call strips out one channel of noise, which is often enough to catch something you would have missed.

Voice coaches and interviewers work from a rough map of what shifts in pitch and pace tend to accompany. Treat it as a starting point:

• Higher pitch than normal: Stress, excitement, or anxiety
• Lower, flatter pitch: Sadness, boredom, or deliberate calm
• Faster speech: High energy (excitement, anxiety, or anger)
• Slower speech: Low energy (sadness, fatigue, or very careful thought)
• Sudden volume drop: Vulnerability, tenderness, or sharing something sensitive

Action Step: The next time you’re on a phone call and want to gauge how someone really feels, close your eyes and focus entirely on their voice. Notice pitch changes, pace shifts, and pauses. You may pick up emotional signals you’d miss face-to-face.

What Video Calls Take Away

Most of the body is gone on video, and what’s left is distorted.

You can’t see hands below the desk, feet, or how someone’s whole torso is angled. Faces sit unusually close and large on screen, which has been proposed as part of why a long call feels more draining than the same conversation in a room¹². And real eye contact is impossible: looking at someone’s face means looking away from your camera, so neither of you ever gets it.

Those proposals have since been tested in a survey of thousands of video-call users, where five separate factors each tracked with how drained people felt. Feeling hemmed into a frame mattered most, then watching your own face, then the sense of being stared at by a wall of people. The effort of producing and reading nonverbal cues came last¹³. Video mostly taxes everything except your people-reading.

So read different things.

• Lean on the voice. The face is close and the body is gone, so pitch and pace matter more than usual.
• Treat pauses carefully. A pause that would seem thoughtful in person can just be lag. Ask before you conclude.
• Hide your self-view. Watching your own face was among the strongest predictors of how drained people felt, and it takes attention off everyone else’s.
• Ask more direct questions. You’ve lost the feet, the posture shifts, and the room. Verbal check-ins are the replacement.

Action Step: In your next video meeting, turn off self-view and drag the speaker’s window directly under your webcam. You’ll read them a little better, and you’ll look more present to them.

More on the mechanics: how to look good on Zoom.

How to Spot Red Flags in the First 5 Minutes

Reading people is also about protecting yourself from people who don’t have your best interests at heart. These are the patterns people who do this for a living watch for early:

• Boundary testing: Asking overly personal questions or making slightly inappropriate comments very early, which can be a way of finding out whether you’ll push back.
• The spotlight shift: Watch whether the person redirects every conversation topic back to themselves within seconds. Consistent self-focus tells you something.
• Comparative compliments: Praising you by putting someone else down (“You’re so much smarter than my last colleague”). Someone who does it about other people in front of you may well do it about you in front of them.
• Incongruence: Words and body language that don’t match, like saying something supportive while the face and posture stay tight. A mismatch is worth more of your attention than any single item above it.

One of these on its own is noise. If you see the pattern again across a few conversations, toxic people covers what to do next.

Action Step: After meeting someone new, ask yourself one question: “Did their words and their body tell the same story?” If the answer is no, treat it as the thing to watch for on the next two occasions you see them.

Why Tone Dies in Text

Most of your working day is text now, and text is the thinnest channel you have.

Readers consistently overrate how much tone they have picked up. They believed they had understood about 89% of the tone in the messages they received. In email they had actually caught 62.8% of it. Read the same words aloud and comprehension climbs to 73%, with the reader no less confident either way¹⁴. Friends did no better than strangers.

Some tones survive the trip better than others. When people wrote short messages aiming for one of eight specific tones, rudeness got through 86% of the time and confidence 81%. Deception managed 59%, and formality 46%¹⁵.

Machines find it hard too. On Google’s own 58,000-comment benchmark, its model did well on emotions with obvious word markers, like gratitude and amusement. It scored close to zero on grief¹⁶, which its authors put down to grief being among the rarest labels in the data and one of the hardest for human annotators to agree on.

Action Step: Before you react to a message that landed badly, write down the two other tones it could have been sent in. If you can’t rule them out from the words alone, that’s your answer: ask, or pick up the phone.

Why Some People Are Harder to Read

If you’ve ever met someone who seems completely “unreadable,” you’re not imagining it. Four things come up again and again in practice:

• Narrow expression range: Some people naturally display fewer facial movements, making their microexpressions harder to catch.
• Trained masking: Poker players, negotiators, and lawyers learn to suppress visible cues.
• You lack baseline data: The less you know someone, the harder it is to spot deviations from their normal behavior. This is why strangers feel harder to read than close friends.
• Different norms: What counts as a normal amount of eye contact, expression, or standing distance is partly learned, and the lessons differ by place. There’s a section further down on what actually varies.

One more factor, and misreading this one does real damage.

Autistic people often use less eye contact and show less facial expressiveness. Both are communication differences. For many people, reduced eye contact is a way of managing sensory input so they can concentrate on what you’re actually saying. And the difficulty is mutual, which is the point of what’s called the double empathy problem: the problem belongs to the pair¹⁷. Asked to judge how autistic people were reacting to staged events they responded to genuinely, non-autistic observers did worse than when judging other non-autistic people, and in three of the four situations they rated the two groups as equally expressive¹⁸.

You can watch it happen. When researchers passed a story down chains of eight people, chains of autistic people and chains of non-autistic people retained about the same amount of detail. The mixed chains lost noticeably more¹⁹. So compare people to how they normally behave. A general template of what engagement looks like will mislead you.

Can you get better at any of this? Yes, if you practice the method itself. Across trials of empathy training for doctors, nurses, and healthcare students, people scored their own empathy higher afterward; the gains an outside observer could see were smaller²⁰. The reviewers judged the training worth doing.

Action Step: Before your next meeting, spend 30 seconds on the other person’s situation. What pressures are they under? What do they want out of this conversation? Walking in with that in mind beats walking in cold.

What Counts as Normal Somewhere Else

Everything above assumes you and the person across from you learned the same rules. Often you didn’t.

Every culture teaches display rules: learned norms about which feelings get shown, to whom, and how strongly. A survey of more than 5,000 people across 32 countries found a universal core, with people across all 32 countries expressing more freely toward their own group, alongside real differences in how much expression is thought appropriate at all²¹.

Eye contact is where this gets garbled most often. The same eye contact does feel different in different places. When Finnish and Japanese adults had a face revealed to them meeting their gaze, their hearts showed the same orienting response, but the Japanese participants rated those faces as angrier and less approachable²². What doesn’t survive measurement is the familiar claim that East Asians simply look less. When researchers strapped eye trackers to pairs of people having actual conversations, the East Asian pairs held more mutual gaze than the Western pairs²³.

Standing distance varies too. Across 8,943 people in 42 countries, the comfortable distance to a stranger averaged about 135 cm, to an acquaintance about 92 cm, and to someone close about 32 cm. Climate explained some of the spread, and not in one direction: people in warmer countries let strangers come closer, while keeping more room around the people closest to them²⁴.

Gestures with fixed meanings diverge across cultures in three separate ways: the same message gets a different gesture, the same gesture means something different, or a gesture exists in one region and nowhere else²⁵.

Cultural differences in communication goes through the workplace signals in detail, and proxemics covers the personal-space zones on their own.

Action Step: When you’re reading someone from a background you don’t share, spend the baseline minutes on a single question: what does comfortable look like for this person? Then compare them only to that.

Three Biases That Distort Your Read

Even with perfect technique, your own brain works against you. Three habits do most of the damage:

1. Confirmation bias. Once you form an impression, you notice everything that agrees with it and skip past everything that doesn’t. Decide someone is untrustworthy and you’ll catch every fidget and miss every straight answer.

2. The halo effect. Notice one good thing (they’re attractive, they’re confident, they gave you a firm handshake) and you’ll quietly assume other good things that have nothing to do with it, like intelligence or honesty.

3. The truth default. We tend to believe people unless something forces us not to, and the 54% hides how lopsided this is: people correctly clear 61% of honest speakers but catch only 47% of liars¹. That default is mostly good for you, and it’s most of why staring harder doesn’t help.

Action Step: Before an important conversation, tell yourself one sentence: “My first impression might be wrong.” It costs nothing and it keeps you open to the evidence that contradicts you.

The average person detects lies at only about 54% accuracy, barely better than flipping a coin.

A Smarter Approach to Lie Detection

Most popular advice about spotting lies is wrong, and it is wrong almost everywhere at once. Asked an open question about how you tell when someone is lying, people across 58 countries named gaze aversion more often than anything else, and it was the top answer in 51 of those countries. Put to a direct question in a second survey spanning 63 countries, 71.5% agreed that liars avoid eye contact²⁶.

It is also a belief people don’t act on. When researchers went back over which behaviors actually moved observers’ verdicts, eye contact barely registered, far below what those same observers said they relied on². It helps to know what you’re actually looking for. In a national survey, 60% of people reported telling no lies in the previous 24 hours, and just over 5% of people accounted for half of all the lies told²⁷. A separate survey of nearly 3,000 people in the UK supported the same pattern²⁸. Most people, most days, are telling you the truth.

So change the task instead.

Lying is mentally expensive. A liar is inventing a story, keeping it consistent, suppressing the truth, and monitoring their own behavior, all at once. Make that work harder and the difference between a liar and a truth-teller gets easier to see.

The Reverse Recall Method: Ask someone to tell their story from the end to the beginning. Truth-tellers manage it, because they’re reading off a real memory. Liars struggle, because they rehearsed forwards. In one study, police officers watching reverse-order interviews identified deception 60% of the time, against 42% for the standard version²⁹.

What interviewers report seeing once the task gets harder:

• Longer pauses and more filler words (“um,” “uh”)
• Simpler language and shorter sentences
• Fewer specific details
• Inconsistencies when you ask the same thing a different way

If deception is the part you care about most, how to spot a liar is the deeper treatment.

Action Step: If something feels off, casually ask the person to walk you through it backwards: “So how did it end up? And what happened right before that?” Watch whether they add rich new detail or go vague and repetitive.

Frequently Asked Questions

What is the best way to read people?

Establish a baseline of how someone normally behaves, then watch for clusters of changes across several channels at once: facial expression, posture, voice, and word choice. No single gesture is reliable on its own, and a change only means something relative to that person’s normal.

Why do I struggle to read people?

Usually it’s one of two things. Either you’re reading single cues where you need clusters, or you’re reading people against a general template when the only useful comparison is their own baseline. Confirmation bias does the rest: once you’ve decided someone is uninterested, you notice every signal that agrees with you.

How can you spot a toxic person in the first 5 minutes?

Watch for clusters of early warning signs: boundary testing (overly personal questions too soon), comparative compliments (praising you by putting someone else down), and a consistent pull of every topic back to themselves. The signal worth most attention is incongruence, when someone’s words and their nonverbal behavior tell different stories.

What does it mean when someone is hard to read?

Some people have a narrow range of facial expression, and some have trained themselves to mask. Others simply use a different nonverbal vocabulary: autistic people, for instance, often use less eye contact and show less facial expressiveness, and neither indicates dishonesty or disinterest. The less baseline data you have on someone, the harder they are to read at all.

Is reading people a real skill?

Yes. Judging whether a stranger is lying, people score around 54%, and when police officers, judges, interrogators, and psychiatrists were tested on the same material as everyone else, they showed no reliable advantage¹. What does improve with practice is the method: building a baseline, reading clusters, and checking your own biases before you act on a read.

Can you read people through their voice alone?

Across five experiments, listeners read emotions slightly more accurately from voice alone than from voice and face together¹¹. A phone call with your eyes closed is the easiest way to test it on yourself.

How to Read People Takeaway

Reading people improves with deliberate practice. Here’s where to put it:

1. Build the baseline first. Give the first two minutes of your next conversation to watching how this person behaves when nothing is at stake.
2. Wait for the cluster. Before you act on a gesture, find two more signals in other channels that agree with it.
3. Read one call with your eyes closed. Pick a phone call this week and listen only for pitch, pace, and pauses.
4. Ask when the channels disagree. If someone’s words and their body tell different stories, that’s your cue to ask a follow-up question.
5. Check your template. Notice whether your read is about this person’s baseline, or about how you expect an engaged person to look.
6. Run the story backwards. When something feels off, ask them to tell it from the end. It costs you one question.

One conversation this week is enough to start.

Want to go deeper on the face? Microexpressions covers the seven expressions in far more detail.