The short answer from decades of research is no: you cannot reliably tell whether someone is lying by watching where their eyes go. The popular idea that liars look in a specific direction, avoid eye contact, or shift their gaze in predictable ways has been tested repeatedly, and the evidence consistently fails to support it. That said, the story is more interesting than a flat “no.” Some involuntary eye-related changes, like pupil dilation and shifts in blink rate, do seem to differ between truthful and deceptive states, and a newer wave of research using precise eye-tracking technology is finding patterns that the naked eye would never catch.
The “Look Up and to the Right” Myth
One of the most persistent claims about eye movements and lying comes from Neuro-Linguistic Programming, or NLP. The idea, which has been taught in law-enforcement seminars, self-help books, and corporate training workshops for decades, is that when a right-handed person looks up and to their right, they are constructing an image (i.e., making something up), while looking up and to the left means they are genuinely remembering. The claim sounds scientific enough that many people accept it uncritically.
Researchers have put this specific hypothesis through rigorous testing, and it falls apart every time. A three-part study published in PLOS ONE found no match between the NLP-predicted eye-movement patterns and what liars and truth-tellers actually did. In one of those experiments, participants who were taught the NLP eye-movement hypothesis beforehand performed no better at detecting lies than a control group who had never heard of it. A third experiment coded eye movements from real high-profile press conferences where people were later confirmed to have been lying or telling the truth, and again, the predicted patterns did not appear.1PubMed Central. The eyes don’t have it: lie detection and Neuro-Linguistic Programming
A separate pair of experiments arrived at the same conclusion. In one, over 200 air passengers were interviewed at an airport, answering some questions truthfully and others deceptively. No reliable relationship emerged between the direction of their eye movements and whether they were lying. In a second experiment where people discussed either their real or a fake occupation, only 3 out of 31 participants showed the eye-movement pattern NLP predicts.2Journal of Police and Criminal Psychology. The direction of deception: neuro-linguistic programming as a lie detection tool At this point, the NLP gaze-direction claim is about as scientifically credible as reading tea leaves.
Does Avoiding Eye Contact Mean Someone Is Lying?
Even people who have never heard of NLP tend to believe that liars avoid eye contact. It is one of the most common folk beliefs about deception, and it is also wrong, at least for adults. A study that analyzed authentic high-stakes liars from real cases found that eye contact was maintained equally during deceptive and truthful statements. The researchers specifically noted that their findings “negate the popular belief amongst both laypersons and professional lie detectors (such as the police) that liars behave nervously by fidgeting and avoiding eye contact.”3PubMed. Suspects, lies, and videotape: an analysis of authentic high-stake liars
There is an interesting age-related wrinkle here, though. A study of children and adults found that 7- and 9-year-olds did maintain significantly less eye contact when lying compared with telling the truth. Seven-year-olds made eye contact about 15% of the time when lying versus 57% when telling the truth. But by ages 11 to 15, and certainly in adults, the difference disappeared. Older children and adults showed essentially the same duration of eye contact whether they were being truthful or deceptive.4PubMed Central. Children’s knowledge of deceptive gaze cues and its relation to their actual lying behavior The likely explanation is straightforward: as we grow up, we learn the social expectation that honest people make eye contact, and liars learn to fake it. By adolescence, people are good enough at managing their gaze that avoidance is no longer a useful signal.
Why Looking Away Is About Thinking, Not Lying
Part of the reason gaze avoidance became associated with dishonesty in the popular imagination is that people do look away when they are working hard mentally. But this happens during any cognitively demanding task, whether you are lying, trying to remember your grandmother’s maiden name, or doing mental arithmetic. Research on what scientists call “gaze aversion” has found that looking away helps suppress environmental distractors when your brain needs to focus inward. One study found that gaze aversion was tied to higher retrieval effort and was rare when memories came to mind spontaneously, suggesting that it kicks in when cognitive effort is needed to access hard-to-reach information, regardless of whether that information is true or fabricated.5PubMed. Why and when do you look away when trying to remember? Gaze aversion as a marker of the attentional switch to the internal world during memory retrieval
This creates a fundamental problem for anyone trying to read deception from eye movements in everyday conversation. A person searching their memory for a genuine but distant detail and a person constructing a plausible lie may both look away, blink less, and appear to be concentrating, because they are. The underlying mental process that drives these behaviors is cognitive effort, and that can accompany truth-telling just as easily as deception.
Cognitive Load and What Happens to Your Eyes
Researchers have tried to get around this problem by deliberately increasing the mental demands of an interview. The logic is that if lying already takes extra brainpower, piling on additional cognitive tasks might amplify the differences between liars and truth-tellers. Some of the findings from this approach involve eye-movement metrics that are too subtle to see without instruments.
One study found that truth-tellers generally had more eye movements during interviews, responded faster, and made fewer inconsistencies, all signs of lower mental workload. Liars, by contrast, showed patterns consistent with higher cognitive load.6Criminal Justice and Behavior. LIE Detection by Inducing Cognitive Load Another study using detailed eye-tracking measurements found that spontaneous lies, those not rehearsed in advance, produced changes in saccades and fixations consistent with greater mental effort compared to truth-telling. But here is the catch: when lies were planned ahead of time, several of those oculometric indicators suggested that the planned lie was actually less cognitively demanding than telling the truth. Blink patterns and pupil measurements indicated that people telling truths sometimes showed greater cognitive load than people delivering well-rehearsed lies.7Applied Cognitive Psychology. Assessing cognitive load using oculometrics to identify deceit during interviews
This finding matters because it means the cognitive-load approach has a built-in vulnerability: it works best for catching people off guard, but rehearsed liars can fly under the radar. Anyone who has prepared a cover story and told it enough times may generate eye-movement patterns that look more relaxed, not less, than a person carefully recounting a complicated truth.
Pupil Dilation and Blink Rate
If the direction of your gaze and how much eye contact you make are unreliable cues, are there other eye-related signals that do shift during deception? Two of the more promising candidates are pupil size and blink rate, both of which are largely involuntary.
Pupil dilation tends to increase slightly when people lie. In one experiment, participants showed greater pupil dilation when delivering deceptive messages than when telling the truth.8PLOS ONE. Truth and lies in your eyes: Pupil dilation of White participants in truthful and deceptive responses to White and Black partners The effect is real but modest, and pupil size also changes with lighting, arousal, emotional state, and a long list of other factors that have nothing to do with honesty, making it unreliable as a standalone indicator.
Blink rate shows an even more nuanced pattern. During the act of lying itself, blink rate tends to drop, consistent with the idea that the brain is allocating resources away from routine functions. Immediately after the lie, though, blink rate rebounds and increases above baseline, a compensatory effect as if the system is catching up on deferred maintenance.9Journal of Nonverbal Behavior. Blinking During and After Lying The pattern is subtle enough that no human observer could track it in real time. It requires high-speed video or specialized sensors. And like pupil dilation, blink rate is sensitive to many variables beyond deception, including fatigue, medication, and even how interesting the conversation is.
Eye Tracking in the Concealed Information Test
Where eye-movement research gets genuinely interesting for deception detection is in structured tests rather than free-form conversations. The Concealed Information Test (CIT) does not ask someone whether they are lying. Instead, it shows them a series of items, only some of which would be meaningful to a guilty person, and measures involuntary recognition responses. Traditionally this was done with skin conductance or brain-wave measurements, but eye tracking has emerged as a viable alternative.
In a mock terrorism experiment, researchers used a paradigm based on involuntary oculomotor inhibition, the way your eye-movement system briefly freezes when it encounters something unexpected or significant. Using high-quality eye tracking, they identified all individually chosen “terror targets” with 100% accuracy based on how participants’ eye movements deviated when those items appeared, and they could also distinguish between guilty and innocent participants with high precision.10Scientific Reports. Concealed information revealed by involuntary eye movements on the fringe of awareness in a mock terror experiment
Other work has found that when people try to conceal recognition of familiar faces, scenes, or objects, their fixation patterns give them away. In one set of experiments, people shown personally familiar faces made fewer fixations and visited fewer areas of the image, whether they were honestly acknowledging recognition or trying to hide it. The effect was strongest for faces and weaker for scenes and objects.11PubMed Central. Eye spy a liar: assessing the utility of eye fixations and confidence judgments for detecting concealed recognition of faces, scenes and objects Newer research is also exploring whether eye-movement patterns that occur before a stimulus even fully appears, anticipatory oculomotor signatures, can distinguish guilty from innocent participants in a CIT.12PubMed. Pre-stimulus oculomotor signatures of deception in the Concealed Information Test
These CIT results are genuinely impressive, but they come with important caveats. The test works because it detects recognition of specific items, not because it reads deception from body language. It can tell whether someone recognizes a crime-scene photo, not whether they are lying about their alibi. And it requires a controlled laboratory setup with precise eye-tracking hardware. You cannot replicate it across a dinner table.
Commercial Eye-Based Lie Detection
Despite the shaky evidence for reading deception from the naked eye, commercial products have already tried to package automated eye-movement analysis as a lie-detection tool. One system that has seen some real-world use is “ocular-motor deception testing,” which uses infrared eye tracking to measure involuntary eye behaviors during structured questioning.
Studies on one such system found that its observed accuracy rates were “not significantly less than published rates of 80%,” which sounds reassuring until you think about what it means practically. An 80% accuracy rate still misclassifies one in five people.13Applied Cognitive Psychology. Ocular‐motor deception testing in civilly detained sexually violent persons: An alternative to post‐conviction sex offender polygraph testing? In a criminal justice context, that error rate is a serious concern. And the tests used in such evaluations are typically structured scenarios that differ considerably from the messy ambiguity of real interrogations or job interviews. The legal system has historically been cautious about admitting lie-detection evidence, and both the scientific and legal communities remain wary of embracing these methods.14Applied Cognitive Psychology. Lie detection: historical, neuropsychiatric and legal dimensions
Culture, Neurodiversity, and the Problem of False Reads
Even if there were a reliable eye-movement cue to deception, using it in the real world would be undermined by vast individual differences. Cultural norms profoundly shape how people use their eyes during conversation. In some cultures, avoiding eye contact with an authority figure is a sign of respect, not dishonesty. Lie-detection strategies built on a single cultural paradigm are inadequate for multicultural settings, and applying them across cultures risks systematic misreadings.15JOALL (Journal of Applied Linguistics and Literature). Model of lie cues detection in cross-cultural communication: insights from linguistic and cultural intersections
Neurodiversity adds another layer of complication. Behaviors like gaze aversion, flat affect, and repetitive movements are commonly believed to be signs of deception, but they are also characteristic of autistic individuals in everyday social situations. Research has found that people are perceived as more deceptive and less credible when they display autistic behaviors, even though the actual association with dishonesty is weak to nonexistent.16Applied Cognitive Psychology. The effect of autistic behaviors on evaluations of deception and credibility in everyday social situations This means that using eye-movement cues to judge honesty doesn’t just fail to catch liars; it actively penalizes people whose natural communication style differs from the norm.
Even within neurotypical populations, personality traits matter. One study found that people with interpersonal psychopathic traits showed abnormal blink-rate dynamics during interviews, with more frequent shifts in blink frequency over short time blocks. These shifts resemble patterns previously associated with dishonesty in laboratory settings.17Personality and Individual Differences. In the blink of an eye: Quantitative blink dynamics predict deceptive personality traits in forensic interviews But the finding cuts both ways: if certain personality profiles produce “deceptive-looking” blink patterns during honest speech, the signal is contaminated at the source.
Why Human Eyes Evolved to Be So Readable
It is worth stepping back and asking why we are even tempted to read eyes for honesty. Humans have evolved strikingly visible eyes compared with other primates. The white sclera that surrounds our irises is unusual in the animal kingdom and makes the direction of our gaze visible from a distance. Experimental work with both human and chimpanzee participants has shown that eyes with uniformly white sclera and a darker iris produce a clear advantage in gaze-discrimination tasks. The white sclera also makes gaze direction robust against natural degradation from factors like shading and distance.18PubMed Central. Experimental evidence that uniformly white sclera enhances the visibility of eye-gaze direction in humans and chimpanzees
This evolved readability almost certainly served cooperative purposes: coordinating attention during hunting, silently directing group members toward threats or resources, and building social bonds through shared gaze. The fact that our eyes broadcast so much information may be precisely why we intuitively believe they should reveal deception. We are wired to read eyes for intention. The mistake is assuming that a system evolved for cooperative signaling doubles as a reliable lie detector. Deception has been under evolutionary pressure for just as long as cooperation, and skilled liars have had millions of years to learn how to hold your gaze while saying something false.
Microsaccades and the Frontier of Involuntary Signals
One area of research that may eventually change the picture involves microsaccades, the tiny involuntary eye movements that happen even when you think your eyes are perfectly still. These movements are modulated by cognitive processes in ways that people cannot consciously control. Research has shown that when people anticipate cognitive conflict, their microsaccade rate drops in a measurable way.19PubMed Central. Anticipation of cognitive conflict is reflected in microsaccades: Evidence from a cued-flanker task Since deception involves a form of cognitive conflict, holding two versions of reality while presenting only one, microsaccade analysis is a natural candidate for future lie-detection research.
The catch, again, is that microsaccades require specialized eye-tracking equipment operating at high temporal resolution. They are invisible to the naked eye by definition. And the leap from “cognitive conflict modulates microsaccade rate” to “we can detect lies via microsaccades” remains a research question, not a proven application. The gap between a statistically reliable group-level effect in a controlled lab and a tool that works on individuals in real-world conditions is enormous, and it is littered with technologies that looked promising in pilot studies but never survived the transition.