For generations we’ve been told to watch the eyes. A liar, the adage goes, can’t meet your gaze. That rule-of-thumb, repeated in boardrooms, classrooms and courtrooms, is now on increasingly shaky ground. A growing body of research shows that eye contact is a poor signal of truthfulness—and, in some situations, the opposite of what we assume.
A near-universal belief meets the data
In 2006, the Global Deception Research Team surveyed people in 75 countries speaking 43 languages. Asked how they detect lies, respondents overwhelmingly pointed to gaze: liars, they said, look away. Approximately two-thirds named gaze aversion as their top cue, a remarkably consistent answer across regions and age groups.
Yet when psychologists pooled the experimental evidence, the signal all but vanished. A widely cited 2003 meta-analysis led by Bella DePaulo in Psychological Bulletin found that the association between looking away and deception had an effect size of 0.03—statistically indistinguishable from zero. A more recent temperature check of the field, a 2023 survey of deception researchers by Timothy Luke and colleagues in Psychology, Crime & Law, reported a rare point of agreement in a contested area: more than 80% of experts said gaze aversion is not a reliable indicator of lying.
“Looking away predicts lying about as well as a coin does.”
Dominance, discomfort and culture complicate the stare
The science suggests eye behaviour tracks many things besides honesty: status signalling, anxiety, cognitive load, and cultural norms among them. A steady stare can be as much about dominance or performance under scrutiny as sincerity. Conversely, people who are shy, neurodivergent or culturally inclined to reduce eye contact with authority may appear evasive even when telling the truth.
That complexity matters in everyday life. We often treat a direct gaze as a proxy for character, rewarding those who meet our eyes and doubting those who look away. The evidence indicates this shortcut is not just imprecise—it can be systematically misleading.
Why the myth endures
Folk rules that feel intuitive survive because they’re easy to remember and occasionally appear to work. Confirmation bias does the rest: we tend to recall the hits and forget the misses. In the case of deception, a listener’s own discomfort can be misattributed to a speaker’s shifty eyes, and confidence in the judgment grows with repetition.
Implications for high-stakes settings
While the research base flags clear limits to eye cues, the belief persists in places where accuracy is critical. Interviews, disciplinary meetings, and investigations sometimes lean on a candidate’s or witness’s gaze as a marker of credibility. The literature summarized above suggests caution. At minimum, practitioners should avoid treating eye contact as a standalone diagnostic, particularly when evaluating people across cultural contexts or under stress.
- Interviews and hiring: Overweighting eye contact risks penalizing qualified but anxious or culturally diverse candidates.
- Education and healthcare: Students or patients who avert gaze may be managing cognitive load or discomfort, not hiding facts.
- Security and investigations: Training that equates a steady gaze with honesty—or looking away with deceit—runs counter to the best available evidence.
What the studies actually show
Deception cues tend to be subtle and variable. No single behavioural marker—least of all gaze—consistently separates truth from lies. Experienced liars can also manage their eye behaviour in anticipation of our expectations, sometimes holding a gaze longer to appear credible. By contrast, truth-tellers preoccupied with being believed may look away while they search memory or regulate nerves.
| Study/Source | Scope | Main finding on gaze |
|---|---|---|
| Global Deception Research Team (2006) | 75 countries; 43 languages | ~2/3 of respondents believed liars look away |
| DePaulo et al., Psychological Bulletin (2003) | Meta-analysis of 100+ studies | Effect size 0.03 for gaze aversion–deception link |
| Luke et al., Psychology, Crime & Law (2023) | Survey of deception researchers | > 80% agreed gaze aversion is not a general diagnostic |
Moving beyond the stare test
Evidence-driven approaches focus less on where someone looks and more on what they say and how accounts change under gentle, specific probing. Consistency with verifiable facts, unexpected details that are hard to fabricate, and the cognitive effort required to maintain a false story are more promising lines of inquiry than eye movements alone.
None of this means eyes tell us nothing. They capture attention, regulate turn-taking, and can signal emotion. But when it comes to detecting deception, the research record points to a simple conclusion: don’t trust the stare.
