What peer-reviewed research, FBI behavioral analysis, and documented field studies actually say about reading nonverbal communication — and what popular claims get wrong. Built for men who need accurate information, not theory.
You have probably heard that 93% of communication is nonverbal — that only 7% of what we say comes through words. This figure comes from UCLA psychology professor Albert Mehrabian's 1967 research and has been repeated so many times it is treated as fact. It is not accurate in the way it is commonly used.
Mehrabian himself has stated publicly that his findings apply only to the communication of feelings and attitudes — specifically like or dislike — and cannot be applied to general communication. His studies used single words recorded on tape, not real conversations. They included only women as subjects. The 7% figure was never meant to suggest that words carry almost no meaning in normal interaction.
Why does this matter? Because body language content — especially in the pickup and dating space — is built on overstated claims. Men are taught that they can read a situation from a glance, that specific gestures mean specific things, that body language is a code you crack. The actual research is more nuanced, more honest, and ultimately more useful.
What the research does confirm, consistently and across multiple studies: nonverbal signals matter significantly in attraction and courtship contexts. Women signal interest and availability primarily through nonverbal behavior. Men systematically misread those signals in predictable ways. And the man who learns to read accurately — without overreading — has a genuine advantage over the man operating on assumption.
This page gives you what the research actually supports. Not what sells.
This is a misapplication of Mehrabian's 1967 research. The actual finding: when someone's words and body language contradict each other in expressing feelings, people tend to believe the body language. That is a real and useful insight. The 93% figure applied to general communication is not what the research shows. Source: Mehrabian, A. (1971). Silent Messages. Wadsworth.
Research published in the Journal of Research in Personality (Haselton, 2003) documented what is called the sexual overperception bias — the consistent tendency for men to interpret female friendliness as sexual or romantic interest when it is not.
A 1987 study by Abbey found that 72% of college women reported that a man had misinterpreted their friendliness as sexual interest. A replication of Haselton's study by Bendixen (2014) in Norway — a more gender-equal culture — confirmed the same pattern held across cultures. The misreading is not random. It is systematic.
The research also found the reverse pattern: men tend to underperceive genuine female interest — missing real signals of attraction when they are present. Men miss women's first courtship glances. They underestimate interest when it is subtle. They overestimate it when a woman is simply warm or friendly.
This creates a double problem. Men approach when they should not — misreading friendliness. And men fail to approach when they should — missing genuine interest. Both errors come from the same source: reading individual signals in isolation rather than reading patterns of behavior in context.
A single signal means almost nothing. A cluster of consistent signals over time tells you something real. Psychologist Monica Moore's 1985 field research at Webster University — observing women in real social settings — documented that women who expressed interest did so through multiple consistent nonverbal behaviors, not single gestures. The man who waits for a pattern before reading interest is the man who reads accurately.
Monica Moore's landmark 1985 study observed women in real social settings — singles bars, university snack bars, libraries — and documented their nonverbal courtship behavior systematically. The findings challenge a common assumption about who initiates.
Women in "mate-relevant" contexts signaled availability and interest far more frequently than women in neutral settings. More importantly — women who signaled more often were approached more often. The frequency of female nonverbal signaling, not her physical attractiveness alone, predicted whether men approached her.
A follow-up finding from ScienceDaily research citing Moore's work: women who signal availability through nonverbal cues are approached more often than women who are more physically attractive but less expressive.
What this means for men: she is often telling you whether to approach before you have said a word. The problem documented in subsequent research is that men miss her first signal. Moore's work noted that men often miss a woman's initial courtship glance — the short darting look that signals openness. By the time he notices her, she may have already signaled multiple times.
A 2004 study published in ScienceDirect confirmed: it is usually the male who makes the first verbal approach — but only after he has received nonverbal signals from the female. Men who approach without those signals are operating on assumption. Men who wait and read the signals are operating on information.
The following signals have been documented in peer-reviewed research as indicators of female interest or openness in social contexts. No single signal is conclusive. Research consistently shows that attraction is communicated through clusters of behavior, not individual gestures. Read patterns, not moments.
Moore (1985) documented what she called the "short darting glance" — brief eye contact followed by immediate eye aversion — as one of the most consistent and commonly missed female courtship signals. Unlike a prolonged stare, the quick glance-and-look-away is deliberate signaling. Research on men in courtship contexts (Crook, 1972, cited in ScienceDirect 2004) found that men are hesitant to approach without substantial eye contact and nonverbal signals of interest.
Do not confuse with: The sustained blank stare of someone who happens to be looking in your direction. The interest signal is a look that makes brief contact and breaks — often with a slight smile or downward look. A woman simply looking around a room is not signaling interest in you specifically.
Research by Paul Ekman — professor emeritus of psychology at UCSF and the world's leading expert on facial expressions — documented the difference between a genuine (Duchenne) smile and a social smile. A genuine smile involves the orbicularis oculi muscle around the eyes — the "crow's feet" area. A social smile does not. Moore (1985) documented genuine smiling as a consistent female courtship signal in field settings. A smile directed specifically at you, involving the eyes, in a context where she has also made eye contact, carries weight. A smile that does not reach the eyes is politeness, not interest.
Do not confuse with: General friendliness. People smile at service workers, neighbors, and strangers in passing. Context determines meaning. A smile in isolation tells you very little. A smile combined with other documented signals tells you something.
Research on proxemics — the study of personal space and body positioning — consistently documents that people orient their bodies toward people and things they are interested in and away from things they are not. Joe Navarro, former FBI counterintelligence agent and author of What Every Body Is Saying, documents that the feet are among the most honest indicators of where a person's interest and attention actually lie — because people consciously control their face and hands more than their feet. If she is in a group but her feet are pointed toward you rather than the group, that is documented as a genuine interest signal.
Do not confuse with: Casual proximity in crowded spaces. Physical closeness forced by environment is not the same as chosen orientation. Assess whether she could orient differently and is choosing not to.
Mirroring — unconsciously matching another person's posture, gestures, or speech patterns — has been documented in research as a sign of rapport and connection. It is largely unconscious, which makes it more reliable than deliberate signals. Allan and Barbara Pease document in The Definitive Book of Body Language that mirroring typically begins appearing when someone is genuinely engaged and interested. If she shifts her posture and you notice her posture has matched yours within the last few minutes — without her appearing to do it deliberately — that is a documented rapport signal.
Do not confuse with: Coincidental similar posture. Mirroring happens over time and in response to you specifically changing position. A one-time posture similarity means nothing.
Moore (1985) documented touch initiation — brushing against someone, a hand on the arm during conversation, a knee touch — as a high-confidence courtship signal. Research on touch in social contexts consistently shows that women in Western cultures are more deliberate and selective about initiating physical contact than men. A woman who initiates touch in a social context where touch is not otherwise expected is expressing something. Grammer et al. (2000) documented touch as one of the clearest signals in the progression of a courtship interaction.
Do not confuse with: Incidental contact in crowded environments or professional contexts where touch norms differ. Touch initiation in a context where it is not environmentally required is meaningful. In a crowded bar where bodies brush constantly, touch means something different than deliberate arm contact in an open space.
Moore (1985) catalogued self-grooming behaviors — hair touching, smoothing clothing, checking appearance — as documented courtship signals in field settings. These behaviors appear to function as self-presentation signals indicating that the person is aware of being observed and is responding to that awareness. They are more significant when they appear specifically in response to making eye contact with you or beginning a conversation than when they occur randomly.
Do not confuse with: General self-maintenance. People touch their hair and adjust their clothes constantly for reasons unrelated to attraction. This signal matters when it appears in direct response to an interaction with you, not in isolation.
This section is more important than the interest signals section. A man who can read genuine disinterest clearly — and respond to it immediately and correctly — is the man who never gets labeled, never overstays his welcome, and maintains his dignity in every interaction regardless of outcome.
The research on misperception (Haselton, 2003) shows that men's errors lean toward overreading interest, not underreading it. The correction for that bias is learning to read disinterest signals with the same precision you apply to interest signals.
Crossed arms, body angled away, shoulders turned, torso rotated to create physical distance — Navarro and Pease both document these as consistent disengagement signals. The body moves away from things it does not want and toward things it does. If her body is orienting away from the interaction — even while she is verbally polite — the body is telling you the truth.
Conversation research documents reciprocal engagement as a key signal of genuine interest. When someone is interested, they contribute to the conversation — they ask questions, they volunteer information, they extend exchanges. Short answers with no reciprocal questions are documented verbal disengagement. Verbal and nonverbal signals together tell a complete story. If her words are minimal and her body is closed, you have your answer.
Navarro documents attention direction as one of the most reliable signals of genuine engagement. Where a person directs their attention is where their interest is. Repeated phone checking during an interaction — beyond a single quick look — signals that their attention is elsewhere. Combined with closed posture and minimal verbal response, this is a clear cluster of disengagement signals.
Eye contact direction is a documented indicator of attention and interest. A person who is genuinely engaged looks at you. A person who is looking past you, scanning the room, or making eye contact with others in the environment is not engaged with the interaction. This is distinct from normal environmental scanning — it is a pattern of eyes moving away from you and not returning.
Thank her politely or simply nod and move. No explanation. No second attempt. No visible reaction. The man who can read disinterest clearly and exit cleanly — without losing composure — demonstrates more genuine confidence than the man who pushed through it. His dignity stays intact. Her impression of him, even in rejection, is often better than he realizes. The interaction is over. Move.
Reading others is half the equation. What you are projecting matters equally — because she is reading you the same way you are reading her, and research consistently shows she is better at it. A 2025 meta-analysis of over 800,000 participants across 1,011 studies confirmed that women outperform men at decoding body language in the vast majority of studies reviewed. She will read you accurately before you have said a word.
Research by Amy Cuddy, Dana Carney, and Andy Yap (2010) at Harvard documented that expansive postures — taking up your natural physical space rather than making yourself small — are associated with confidence and are perceived as such by observers. Navarro documents that people who are genuinely comfortable with themselves take up their natural space without aggression or performance. Slouching, hunching, or making yourself physically smaller communicates insecurity more reliably than any verbal attempt to compensate for it can correct.
Research on eye contact in social interactions distinguishes between confident eye contact — steady, warm, present — and the kind of sustained staring that registers as threatening or aggressive. The Haselton (2003) research showing men overperceive interest is partly explained by men holding eye contact longer than the interaction warrants. Navarro documents that comfortable, natural eye contact — making contact, holding briefly, looking away naturally, returning — projects confidence without aggression. Staring without looking away is documented as a dominance or threat signal, not an attraction signal.
Navarro documents excessive fidgeting — leg bouncing, hand wringing, excessive touching of the face or neck — as consistent indicators of anxiety and discomfort that observers read accurately and unconsciously. A man who is genuinely at ease with himself does not move excessively without purpose. Stillness is not rigidity — it is the natural physical state of someone who is not fighting internal anxiety. It projects confidence more reliably than any deliberate gesture.
Pease documents that people who are anxious or seeking approval tend to move and speak faster than the interaction requires — rushing to fill silence, moving quickly, speaking at a pace that outpaces the other person's comfort. Deliberate pacing — moving at a natural speed, comfortable with silence, not rushing to fill gaps — is consistently read as confidence. It cannot be faked without practice because it requires genuine comfort, not performance of comfort.
Pease addresses this directly. Crossed arms can indicate defensiveness — but they can also indicate that someone is cold, physically comfortable in that position, or simply their habitual resting posture. A single gesture read in isolation is unreliable. Context and clusters of signals are what tell you something meaningful. A woman with crossed arms who is making eye contact, leaning in, and engaged in the conversation is not closed off.
Pupil dilation does occur in response to emotional arousal including attraction — this is documented. However, pupils also dilate in low-light environments, in response to certain medications, and in response to any emotionally arousing stimulus including fear. You cannot reliably read pupil dilation in normal social lighting conditions, and attempting to do so is not a practical or accurate approach to reading interest.
Moore (1985) did document self-grooming including hair touching as a courtship signal in field settings. However, this only carries meaning when it appears in direct response to an interaction with you specifically — not as a general behavior pattern. Many people touch their hair constantly as a habitual behavior unrelated to any interpersonal dynamic. Read the context: is she touching her hair after making eye contact with you, or is she doing it while looking at her phone?
This claim is documented as one of the most persistent and least supported in body language research. Navarro — whose entire career was built on behavioral analysis for the FBI — states explicitly that there is no single reliable indicator of deception. Research by Ekman and colleagues found that most people, including trained law enforcement, perform at near chance levels when trying to detect lies from body language alone. The idea that you can spot a liar from crossed arms or avoiding eye contact is not supported by the evidence. This page does not include deception signals because the research does not support presenting them as reliable.
Every signal on this page must be read inside its context. A coffee shop reads differently from a bar. A first conversation reads differently from a third date. A professional setting reads differently from a social one. A woman who is naturally warm and expressive reads differently from a woman who is naturally reserved.
The most dangerous body language mistake a man can make is applying a fixed interpretation to a signal without accounting for context. Research on the sexual overperception bias (Haselton, 2003) found that men's misperceptions were highest in environments with high sexual relevance — bars, parties — where the context itself primes the interpretation of ambiguous signals as interest. The same behavior in a different environment means something different.
Before reading her, read the environment. Before reading the environment, read whether the moment is appropriate for an interaction at all. This is the order that produces accurate reading — and it is the order that produces respect for another person's context, which is ultimately what separates a man with genuine situational awareness from one who is simply executing a checklist.
One signal means nothing. Three or more consistent signals pointing in the same direction — that is information. Before you decide the interaction is going somewhere, count how many documented interest signals are present simultaneously. If you can only identify one, you do not have enough data.
Moore's research shows women signal availability before men approach. Before you approach someone, observe whether she has done anything documented — a glance and look away, open body orientation, a genuine smile in your direction. If she has, you are moving on information. If she has not, you are moving on assumption. Both are legitimate — but know which one you are doing.
The research shows men systematically overread friendliness as interest. When you feel certain someone is interested, apply a deliberate correction: am I reading genuine signals, or am I reading warmth? Are her signals specific to me, or is she warm with everyone in this environment? This is not self-doubt — it is accuracy.
The moment you register closed posture, minimal verbal engagement, diverted attention, and body orientation away from you — accept the information and exit cleanly. No second attempt. No explanation. A nod and a move. The man who does this correctly is not defeated. He is operating with accurate information and responding to it appropriately.
Before you focus on reading others, focus on what you are projecting. Natural posture that takes up your genuine space. Eye contact that is present without being a stare. Stillness that comes from actual composure, not performance of it. Pace that is unhurried. These cannot be faked reliably — they are built through genuine comfort with who you are. That is the work that precedes everything on this page.