Is beauty purely subjective, or do recurring visual cues shape ratings? Research finds associations with proportions, averageness, symmetry, skin appearance, and other features, while culture, the rater, the image, and the study design still matter. No single geometric formula defines an objectively beautiful face.

The Golden Ratio (Phi) in Facial Beauty

The golden ratio — approximately 1.618, denoted by phi (φ) — is often promoted as a facial ideal. The scientific evidence is more limited. Schmid, Marx, and Samal (2008) developed a geometry-based attractiveness index incorporating neoclassical canons, symmetry measures, and golden-ratio terms. Their model showed that facial geometry could be associated with ratings in their data; it did not establish that every attractive face follows phi or validate a universal ideal template.

Pallett, Link, and Lee (2010) used controlled image manipulations and found peak ratings near two mean-face proportions:

  • The vertical distance from the eyes to the mouth was about 36% of face length
  • The distance between the eyes was about 46% of face width

The authors called these new "golden" ratios because they produced optima in the experiment, not because they equal φ. The values matched the average proportions of the faces studied and should not be converted into unsupported phi-based rules for the nose, lips, jaw, or whole face.

The Averageness Hypothesis

One of the most robust findings in attractiveness research is that average faces are attractive faces. This doesn't mean plain — it means faces with proportions close to the mathematical average of a population.

Langlois and Roggman (1990) pioneered this research by creating computer-averaged composite faces. They found that composites of 16 or 32 faces were rated as significantly more attractive than any individual face. This has been replicated across many cultures, supporting the notion that averageness signals genetic diversity and health.

Symmetry: The Universal Attractor

As discussed in our article on facial symmetry and success, bilateral symmetry is one of the strongest cross-cultural predictors of attractiveness. Perrett et al. (1999) demonstrated that symmetrized versions of faces are consistently preferred over original, asymmetric versions.

Sexual Dimorphism

Faces that display sex-typical features — masculine features in men and feminine features in women — are generally rated as more attractive, though with important caveats.

Little et al. (2011) in their comprehensive review in Philosophical Transactions of the Royal Society B noted:

  • Feminine facial cues in women are often preferred in the samples reviewed, with variation across contexts
  • For men, preferences for masculinity vary by context — moderate masculinity is often preferred over extreme
  • Hormonal cues (jaw width, brow ridge) signal genetic fitness

Skin Quality

Often overlooked, skin appearance is a major determinant of attractiveness. Research by Jones et al. (2004) showed that skin color homogeneity — evenness of tone — is a stronger predictor of perceived health and attractiveness than facial symmetry in some contexts.

Fink et al. (2006) confirmed that skin texture and color distribution influence attractiveness ratings independently of facial structure, highlighting the importance of skincare in overall appearance.

Cross-Cultural Consistency

A critical question is how much these patterns generalize. Langlois et al. (2000) reported substantial agreement across samples, alongside meaningful cultural and methodological variation. Symmetry, averageness, skin appearance, and sexual dimorphism may contribute to ratings, but they do not establish a single universal beauty standard.

What AI Face Analysis Measures

Modern AI face rating technology can use facial landmarks and model-specific reference patterns to produce structured assessments. These outputs can help compare photos, but they are estimates rather than objective measures. Depending on the system, they may evaluate:

  • Facial proportions relative to predefined or data-derived reference patterns
  • Symmetry scores across multiple facial regions
  • Feature harmony and balance
  • Skin quality indicators

Key Research References

  • Schmid, K., Marx, D., & Samal, A. (2008). "Computation of a Face Attractiveness Index Based on Neoclassical Canons, Symmetry, and Golden Ratios." Pattern Recognition, 41(8), 2710–2717. https://doi.org/10.1016/j.patcog.2007.11.022
  • Pallett, P.M., Link, S., & Lee, K. (2010). "New 'Golden' Ratios for Facial Beauty." Vision Research, 50(2), 149–154. https://doi.org/10.1016/j.visres.2009.11.003
  • Langlois, J.H. & Roggman, L.A. (1990). "Attractive Faces Are Only Average." Psychological Science, 1(2), 115–121. https://doi.org/10.1111/j.1467-9280.1990.tb00079.x
  • Little, A.C., Jones, B.C., & DeBruine, L.M. (2011). "Facial attractiveness: evolutionary based research." Phil. Trans. R. Soc. B, 366, 1638–1659.
  • Jones, B.C. et al. (2004). "Facial skin coloration affects perceived health." International Journal of Primatology, 25(6), 1213–1228.
  • Fink, B., Grammer, K., & Matts, P.J. (2006). "Visible skin color distribution plays a role in the perception of age, attractiveness, and health." Evolution and Human Behavior, 27(6), 433–442.
  • Langlois, J.H. et al. (2000). "Maxims or Myths of Beauty?" Psychological Bulletin, 126(3), 390–423.