What Makes a Face Attractive? The Science Explained

What makes a face attractive? It's a question humans have pondered for millennia, from ancient Greek sculptors chiseling idealized proportions to modern psychologists running eye-tracking experiments. While beauty may feel deeply personal and subjective, decades of research reveal that certain features consistently drive attractiveness judgments — often across cultures, ages, and even species boundaries. The science of attractiveness turns out to be far more universal, and far more interesting, than most people expect.

Symmetry: The Foundation of Facial Appeal

Of all the factors that influence facial attractiveness, symmetry is arguably the most well-established. Gillian Rhodes, a psychologist at the University of Western Australia, conducted a landmark meta-analysis in 2006 examining dozens of studies on symmetry and attractiveness. Her conclusion was clear: across a wide range of experimental designs and populations, more symmetrical faces are consistently rated as more attractive.

But why should symmetry matter? The leading explanation comes from evolutionary biology. Symmetry is thought to be an honest signal of developmental stability — the ability of an organism to develop normally despite environmental stresses like pathogens, toxins, and nutritional deficiencies during growth. Small random deviations from perfect symmetry, known as fluctuating asymmetry, accumulate when development is disrupted. A highly symmetrical face, then, may advertise a robust immune system and strong genes.

David Perrett and his colleagues at the University of St Andrews have explored this further using digitally manipulated faces. In a series of studies, they created symmetrical versions of participants' faces by blending each half with its mirror image. Participants consistently preferred the symmetrical versions, even when the differences were subtle enough that they couldn't consciously identify what had changed. This suggests that our sensitivity to symmetry operates partly below conscious awareness — a rapid, intuitive assessment rather than a deliberate calculation.

It's worth noting that perfect bilateral symmetry is virtually nonexistent in real human faces. Everyone has minor asymmetries. What matters for attractiveness isn't perfection but rather the degree of deviation. Faces closer to symmetrical tend to score higher, but the relationship is a gradient, not a binary threshold.

Proportions and the Golden Ratio

Beyond symmetry, the spatial relationships between facial features play a major role in attractiveness. Researchers have identified several proportional relationships that correlate with higher attractiveness ratings. Among the most discussed is the golden ratio — the mathematical constant phi (approximately 1.618) that appears throughout nature, from nautilus shells to spiral galaxies.

In 2010, a research team led by Pamela Pallett at the University of California, San Diego, identified specific "ideal" facial proportions. They found that faces were rated most attractive when the vertical distance between the eyes and mouth was approximately 36% of the face's total length, and when the horizontal distance between the eyes was about 46% of the face's width. These proportions happen to align closely with the ratios produced by averaging many faces together — an important connection we'll explore shortly.

Several key proportional relationships have been linked to attractiveness ratings in empirical studies:

• The ratio of face width to face height
• The spacing between the eyes relative to face width
• The vertical position of the mouth between the nose and chin
• The relationship between forehead height, nose length, and lower face height (the "facial thirds")

While the golden ratio makes for compelling headlines, the scientific picture is more nuanced than popular accounts suggest. Not every attractive face conforms to phi, and not every face that matches phi is considered attractive. Proportions matter, but they interact with many other features. We explore this topic in much greater depth in our dedicated article on the golden ratio in faces.

The Averageness Hypothesis

One of the most counterintuitive findings in attractiveness research is that average faces are highly attractive. This doesn't mean "plain" or "mediocre" — it means faces whose features are close to the mathematical mean of a population.

The foundational study was published in 1990 by Judith Langlois and Lori Roggman at the University of Texas at Austin. They digitally blended photographs of faces together to create composite images. As more faces were averaged into the composite, the resulting image was rated as increasingly attractive. A composite of 32 faces was rated significantly more attractive than the vast majority of individual faces that went into it.

Why would this be? There are several complementary explanations. First, averaging smooths out asymmetries and unusual proportions, producing the kind of symmetrical, well-proportioned face we've already discussed. Second, average features may signal genetic diversity. A face that sits near the population center is less likely to carry homozygous (double-copy) recessive genes associated with developmental abnormalities. Third, from a cognitive processing standpoint, average faces may simply be easier for the brain to process — they match our stored mental prototype of "face" more closely, creating a fluency effect that we experience as aesthetic pleasure.

Importantly, average does not mean maximally attractive. Research by Perrett and others has shown that the most attractive faces actually deviate from the average in specific, predictable ways — exaggerated femininity in female faces and certain masculinized features in male faces. Averageness provides a strong baseline of attractiveness, but the peak lies slightly beyond the mean.

Sexual Dimorphism

Sexual dimorphism — the degree to which a face displays typically masculine or feminine characteristics — is another powerful driver of attractiveness, though its effects are more complex and context-dependent than symmetry or averageness.

Masculine features in male faces include a prominent brow ridge, a wider and more angular jaw, a broader chin, and thinner lips. These features are shaped in part by testosterone exposure during puberty. A strong jawline and prominent brow signal high testosterone levels, which are associated with physical strength and immune function — but also with aggression and lower parental investment.

Feminine features in female faces include larger eyes relative to face size, fuller lips, a smaller nose, higher cheekbones, a more tapered chin, and smoother skin. These features are associated with estrogen, which is linked to fertility and youth. Across cultures, feminized female faces are consistently rated as more attractive.

For male faces, the picture is more complicated. While some studies find that women prefer masculine faces, others find a preference for slightly feminized male faces. Research by Anthony Little, Benedict Jones, and colleagues has shown that these preferences shift with context. Women tend to prefer more masculine faces when evaluating short-term partners or during the fertile phase of the menstrual cycle, but prefer slightly less masculine faces when evaluating long-term partners. This pattern has been interpreted as a trade-off between "good genes" (signaled by masculinity) and "good dad" qualities (signaled by softer, more approachable features).

Environmental factors matter too. Studies across different countries suggest that women in environments with higher pathogen prevalence show stronger preferences for masculine faces, possibly because the immune benefits signaled by testosterone become more important when disease risk is high.

Skin Quality as a Health Signal

While symmetry and proportions receive the most public attention, skin quality may be one of the most potent drivers of attractiveness — and one of the most underappreciated. Research by Bernhard Fink, Karl Grammer, and their colleagues has consistently found that skin texture, color homogeneity, and luminance distribution are strong predictors of attractiveness ratings, sometimes outperforming structural features like symmetry.

Skin quality functions as a health and age signal. Smooth, evenly pigmented skin with consistent coloring suggests youth, good nutrition, absence of parasitic infection, and hormonal health. Conversely, blotchiness, rough texture, and uneven pigmentation are associated with aging, sun damage, illness, and stress. Our visual system appears to be remarkably sensitive to these cues, even when we're not consciously evaluating skin.

Several specific skin attributes have been linked to attractiveness in controlled experiments:

• Color homogeneity — even skin tone without patches of redness, hyperpigmentation, or sallowness
• Texture smoothness — absence of visible pores, roughness, or blemishes
• Apparent luminance — healthy skin reflects light in a characteristic way that differs from unhealthy skin
• Color distribution — slight rosiness (indicating oxygenated blood flow) and mild yellowness (indicating carotenoid intake from fruits and vegetables) are both associated with perceived health

Ian Stephen and colleagues at the University of St Andrews demonstrated in 2009 that increasing skin yellowness and redness in facial photographs made faces look healthier and more attractive. The carotenoid pigments responsible for skin yellowness come from fruit and vegetable consumption, meaning that diet literally shows up on your face — and people find it attractive.

Across cultures, skin quality is one of the most consistent predictors of attractiveness ratings. While preferences for specific skin tones vary across populations, the preference for even, clear, smooth skin is nearly universal. You can measure your own facial features, including skin quality indicators, with our free analysis tool.

Cultural Universals vs. Variations

One of the most important questions in attractiveness research is the extent to which beauty standards are universal versus culturally constructed. The evidence points firmly toward a hybrid answer: some aspects of attractiveness are remarkably consistent across human populations, while others show meaningful cultural variation.

Cross-cultural universals include:

• Preferences for facial symmetry
• The attractiveness of average (population-mean) facial configurations
• Preferences for clear, even skin
• The attractiveness of feminized features in female faces
• Preferences for facial features signaling youth in female faces

Langlois and colleagues published a major meta-analysis in 2000 reviewing 919 studies on attractiveness. They found strong cross-cultural agreement in attractiveness judgments, including between Western and non-Western populations, and even in studies involving infants too young to have absorbed cultural beauty standards. Infants as young as two to three months old gaze longer at faces that adults rate as attractive — a finding that is difficult to explain through cultural learning alone.

Cultural variations tend to cluster around features that are less directly tied to health signaling:

• Body weight and facial adiposity preferences — societies with greater food scarcity tend to prefer heavier faces and bodies, while resource-abundant societies prefer thinner features. This was demonstrated in work by Viren Swami and Martin Tovée across multiple countries.
• Skin tone preferences — while even skin is universally preferred, the ideal shade varies. Many cultures show a preference for lighter skin within their population range, but this pattern is far from universal and is heavily influenced by historical and socioeconomic factors.
• Specific feature preferences — the ideal nose shape, lip fullness, and eye shape show more variability across populations, likely reflecting familiarity effects and local adaptation.
• Masculinity preferences in male faces — as discussed above, these vary with ecological context, pathogen prevalence, and resource availability.

The evolutionary psychologist Anthony Little has argued that the universal components of attractiveness reflect adaptations for mate quality assessment — detecting symmetry, health, fertility, and genetic compatibility. The variable components, meanwhile, reflect calibration to local conditions: what signals status, health, or fertility in one environment may differ from another.

Putting It All Together

What makes a face attractive, then, is not any single feature but a convergence of signals. Symmetry advertises developmental stability. Proportions near the population average suggest genetic diversity and make faces easy to process. Sexual dimorphism signals hormonal health and reproductive fitness. And skin quality provides a real-time readout of current health, nutrition, and youth.

These signals interact in complex ways. A face with excellent symmetry but poor skin quality won't rate as highly as one with both. Extreme masculinity might be attractive in one context and off-putting in another. And individual preferences — shaped by personal experience, familiarity, and cultural context — add another layer of variability on top of the universal foundations.

Perhaps the most reassuring finding from decades of attractiveness research is that beauty is genuinely multidimensional. There is no single "most attractive" face template. Instead, there are many paths to attractiveness, and features like skin health and grooming — which are at least partly under our control — matter just as much as the bone structure we're born with.