Facial symmetry and attractiveness: what the evidence actually supports

Looksmaxxing rests on a premise worth examining directly: that there is an objective, measurable ideal face — highly symmetrical, "average" in its proportions, strongly masculine or feminine — and that moving toward it maximises attractiveness and even signals underlying health. The three ideas are real hypotheses in the science of face perception. But the evidence that facial features are honest indicators of health is decidedly mixed, and a recent review of the whole question reached no firm conclusion at all [s1]. The forums treat the ideal as settled fact. The research does not.

The "honest signal" hypothesis, tested

The idea underpinning much of this is that a symmetrical, average, sex-typical face advertises "developmental stability" — that its owner grew up buffered against genetic and environmental stress, and is therefore healthier. It is an appealing story, and it is why attractiveness is so often described in the popular telling as a proxy for good genes. A 2025 scoping review set out to test it, searching the major databases for studies linking facial features — specifically symmetry, averageness and sexual dimorphism — to actual health [s1]. Of 702 articles screened, 21 met the criteria [s1].

What those studies examined was broad: associations between facial appearance and cardiovascular health, immune function, oxidative-stress levels, cortisol, reproductive health, cognitive health and general physical health [s1]. What they found, taken together, did not settle the question. The review's conclusion is that the evidence offers "an inconclusive answer" to whether facial features can serve as honest indicators of health, and it explicitly warns against using the face as a biomarker of someone's biological condition [s1]. In other words, the load-bearing assumption of looksmaxxing — that optimising your face optimises a readout of your health — is not something the literature supports.

That does not mean symmetry and averageness have no bearing on rated attractiveness; they are among the most-studied cues in the field, and people do, on average, prefer them. It means the further claim — that these features are a window onto health, and therefore worth extreme measures to maximise — is where the science thins out to nothing conclusive.

The ideal is partly manufactured

There is a second problem with the "measurable ideal," which is that the version many people are now measuring themselves against is synthetic. A 2025 study quantified what TikTok's "Bold Glamour" filter does to a face, using facial-landmark software on images of 10 participants before and after [s2]. The filter sharpened the nose, fuller-ed the lips, lifted a brow, raised the cheekbones and increased jaw prominence — and, tellingly, it "enhanc[ed] symmetry" [s2]. The individual changes were mostly under 10%, small enough to look natural while adding up to a markedly more "ideal" face [s2]. The authors flagged the obvious risk: unrealistic beauty standards and the toll on psychological well-being [s2].

So the symmetrical, high-cheekboned template circulating as the target is, in part, an algorithm's output — a face engineered past what unedited faces show, presented as the baseline. Chasing it is chasing a moving, artificial line.

What the forums do with a mixed literature

The looksmaxxing communities take this uncertain, hedged science and treat it as a precise specification. The sociological study of a looksmaxxing forum with 6 million monthly visitors, drawn from 8,072 comments, documented members dissecting one another's faces for exactly the features the research treats as weak, ambiguous cues — brow ridges, jaw angles, small asymmetries — and ranking each other on them [s3]. The gap between "symmetry is one modest, uncertain correlate of rated attractiveness" and "your slightly uneven face is a fixable defect that defines your worth" is the gap the subculture manufactures, and it is where the harm begins. That preoccupation with a minor or imagined flaw is also the defining feature of body dysmorphic disorder, covered in our report on looksmaxxing and body dysmorphic disorder.

The bottom line

Symmetry, averageness and sexual dimorphism are genuine topics in the study of attractiveness, and faces closer to those templates do tend to be rated more attractive on average. But the stronger claim the trend depends on — that the ideal face is an honest signal of health worth extreme effort to attain — is not supported; the most recent review of the evidence calls it inconclusive and cautions against reading health off a face at all [s1]. Meanwhile the template itself is increasingly set by filters that outrun any real face [s2]. The reasonable reading is that a small, uncertain effect has been inflated into a specification, and then into a source of distress. The practical looksmaxxing habits with real evidence behind them sit elsewhere — in skin and sun protection — and none of them require believing your bone structure is a health certificate.

This article is informational and is not medical advice.

Sources

• [s1] Is the human face a biomarker of health? – A scoping review — PLOS One, 25 August 2025. https://doi.org/10.1371/journal.pone.0318138
• [s2] Artificial Intelligence Beauty Filters and Aesthetic Surgery: Insights from TikTok's Bold Glamour Filter — Plastic and Reconstructive Surgery – Global Open, 27 October 2025. https://doi.org/10.1097/gox.0000000000007133
• [s3] When Help Is Harm: Health, Lookism and Self-Improvement in the Manosphere — Sociology of Health & Illness, 11 March 2025. https://doi.org/10.1111/1467-9566.70015