Facial Width-to-Height Ratio and Behavior: What the Evidence Actually Shows

For a while now, a simple facial ratio has been treated as a kind of shorthand for character. The claim goes like this: people with wider, shorter faces are more aggressive, more dominant, less trustworthy. It’s a tidy story, and tidy stories about faces tend to spread fast. The trouble is that the more rigorously anyone has tested it, the less of it survives.

⸻

What the Ratio Actually Measures

Facial width-to-height ratio (fWHR) is calculated by dividing bizygomatic width — the distance between the cheekbones — by upper facial height, typically measured from the upper lip to the mid-brow (some studies use nasion-to-brow variants instead). That measurement choice matters more than it sounds like it should: manual landmarking, automated tools like Face++, and different height definitions don’t always agree with each other, which is one of the quieter reasons early findings didn’t hold up as well as expected.

⸻

Where the Idea Came From

The original hypothesis grew out of several small, lab-based studies from the 2010s that reported wider-faced men scoring higher on measures of aggression and dominance in narrow, specific samples. It was an appealing idea — a visible trait doing the work of a personality test — and it got picked up quickly, the way findings from small samples sometimes do before anyone’s had the chance to stress-test them. It’s the same pattern that shows up whenever a single trait gets asked to explain a lot of human behavior, something I got into when unpacking Cialdini’s work on persuasion and the shortcuts our judgment relies on.

⸻

The Large Studies That Complicated the Story

The correction came from scale. Kosinski (2017), published in Psychological Science, drew on roughly 137,000 participants and over 170,000 images from the myPersonality dataset and tested fWHR against 55 separate psychometric scales. The correlations were negligible. It’s worth being precise about the caveats here too — the personality measures were self-reported, the sample was drawn from Facebook-linked profiles rather than a random cross-section of the population, and facial measurements combined automated and manual methods. None of that erases the finding, but it does mean “no substantial link” is the accurate phrase, not “no link” full stop.

A second large study, Wang, Nair, Kouchaki, and colleagues (2019), used evaluator ratings and executive samples and reached a similar conclusion, proposing that earlier associations may have reflected biased perception — the way people expect a wider face to signal dominance — rather than any real behavioral difference underneath. A 2018 study by Zhang and Hahn in PLOS ONE found no association between fWHR and women’s sexual desire either, adding to a pattern of large, well-powered studies coming up empty where small ones had reported effects. As for the personality-inventory research some early posts on this topic cite — the specific 16PF findings are hard to pin down to a single robust study, and where they exist, the correlations reported were weak and inconsistent across gender, not a basis for treating fWHR as predictive of anything.

⸻

Why the Findings Diverge

Small samples, narrow populations, and inconsistent facial-measurement techniques go a long way toward explaining why early work found effects that didn’t replicate at scale. This is a familiar shape in behavioral science — it’s the same reason a missed deadline gets read as carelessness by one manager and bad luck by another, something I looked at in how snap judgments about behavior get formed and hardened. Small, uncontrolled samples tend to manufacture the exact kind of pattern our judgment is already primed to find.

⸻

What This Means Beyond the Study

The practical concern isn’t academic. Treating facial structure as a stand-in for personality has a long, uncomfortable history, and it resurfaces any time a new metric gets marketed as objective. Bias against people who simply look a certain way isn’t hypothetical — it’s the same mechanism that shows up in how introverts get read as cold or unfriendly, a stereotype I’ve written about in why quiet doesn’t mean unkind. A ratio on someone’s face carries no more information about who they are than their preferred seating at a party does, and the evidence increasingly says exactly that.

⸻

References

• Kosinski, M. (2017). Facial width-to-height ratio does not predict self-reported behavioral tendencies. Psychological Science.
• Wang, D., Nair, K., Kouchaki, M., et al. (2019). A case of evolutionary mismatch? Why facial width-to-height ratio may not predict behavioral tendencies. Psychological Science.
• Zhang, Y., & Hahn, A. C. (2018). No evidence that facial width-to-height ratio (fWHR) is associated with women’s sexual desire. PLOS ONE.
• Wikipedia contributors. (2025). Facial width to height ratio. Wikipedia.

⸻

Further Reading

• Facial width-height ratio does not predict self-reported behavioral tendencies — Stanford GSB faculty summary.
• Facial width-to-height ratio does not predict self-reported behavioral tendencies — review and methods breakdown.
• Systematic review of fWHR and personality/behavior correlates.

⸻

Note: This article is for informational purposes only and does not constitute professional advice. Always consult with a qualified professional for specific guidance.