Midface · Three-Axis Score

Cheekbone test

RealSmile Research Team · Facial Analysis Specialists

Updated May 19, 2026

Based on 5 peer-reviewed sources

Width, projection, hollow depth. Three measurable cheekbone axes scored against published anthropometric norms.

High cheekbones are not one thing. They are three things that have to combine. The scan tells you which of the three your face has and which is the bottleneck.

17 metrics · Free · No signup

Score My Cheekbones →

Free score · $14.99 unlocks the written midface report

The three axes the scan measures

Bizygomatic width index. Horizontal distance between the two zygion landmarks divided by total face width. Faces read as high-cheekboned sit at roughly 0.75 to 0.80 on this index, per Farkas anthropometric norms. Below 0.70 reads as narrow; above 0.85 reads as wide-set.

Malar projection index. Perpendicular distance from the malar apex to the maxilla plane, normalized by midface length. Measured from a three-quarter angle. The reference range is 0.18 to 0.24 of midface length for faces consistently rated as having projecting cheekbones.

Submalar hollow contrast. The luminance and depth differential between the malar apex and the soft tissue immediately below. Deeper hollows produce stronger visible cheekbone read even at constant width and projection. This is the axis that responds most to soft-tissue changes.

The composite cheekbone score is the rank-normalized combination of the three axes against published norms. The 17-metric scan returns the composite percentile and the three sub-scores so you can see exactly which axis is high and which is low.

What changes which axis

Bizygomatic width

Does not change in adults. Bone structure is fixed after late adolescence. Width can only be visually narrowed (lower-face shaping) or framed (haircut, brow shape) to alter the perceived ratio.

Malar projection

Does not change in adults without surgery. Filler or implants can add projection. Soft-tissue volume loss around the cheekbone reveals existing projection that was previously hidden by fullness.

Submalar hollow depth

Responds to body composition. A 3 to 5 percent body fat reduction visibly deepens the hollow. Masseter reduction (dietary, postural, or pharmacological) narrows the lower face and increases the contrast.

Perceived width ratio

Responds to framing. A wider haircut, fuller brow, or narrower beard line changes the ratio of cheekbone width to head outline width without changing the bone.

Perceived projection

Responds to lighting and angle. A three-quarter angle at 30 to 45 degrees with forward light produces the highest perceived projection from the same underlying bone.

Perceived cheekbone score

Composite of all of the above. The free scan returns the composite percentile. The $14.99 report ranks the levers by impact on your specific weak axis.

Why the three axes matter separately

Most people asking "are my cheekbones high?" are conflating the three axes. A face can be high on bizygomatic width and low on projection (wide, flat midface). A face can be high on projection and low on width (narrow, defined midface). A face can be high on both and low on hollow contrast (full, framed cheekbones that do not photograph as gaunt).

The configuration matters for what changes are worth pursuing. If your projection is high but hollow contrast is low, body composition reduction has a real payoff. If your projection is low at any width, no body composition change will produce a Hollywood-cheekbone look — the bone is not there. The scan separates the diagnosis from the wish.

Honest limits

Bone structure does not change after late adolescence. No diet, exercise, or facial massage moves bizygomatic width or malar projection.
Hollow depth responds to body composition but with diminishing returns. Below roughly 12 percent body fat in men or 18 percent in women, further reduction starts to read as gaunt rather than defined.
Photo angle and lighting change perceived projection by 20 to 40 percent. Casual mirror comparisons will mislead you about real change. Use the scan, which normalizes for head pose.
The published attractiveness effect of cheekbone prominence is moderate and non-linear (Perrett et al., 1998). High prominence is desirable up to a point; extreme prominence reduces ratings.
Cheekbone prominence interacts with other midface metrics. A high cheekbone score does not predict overall facial attractiveness on its own — the full 17-metric report shows the context.

Cheekbone test FAQ

What makes cheekbones look high?+

Three measurable inputs combine to produce the visible "high cheekbone" impression. First, bizygomatic width relative to face width — the cheekbone width sits at roughly 75 to 80 percent of head width in faces read as high-cheekboned. Second, malar projection forward of the maxilla — the cheekbone sits proud of the surrounding tissue rather than flush with it. Third, submalar hollow depth — the soft tissue below the cheekbone recedes visibly, creating shadow that emphasizes the bone above. Faces lacking one of the three rarely read as high-cheekboned even if the other two are present.

How are cheekbones actually measured on a photo?+

A frontal photo measures bizygomatic width as the horizontal distance between the two zygion landmarks (the most lateral point of each zygomatic arch). A three-quarter or profile photo measures malar projection as the perpendicular distance from the cheek apex to the maxilla plane. Submalar hollow depth measures the relative recession of the soft tissue below the cheekbone versus the malar apex. The 17-metric scan computes all three from a single photo and returns them as a normalized score against population means published by Farkas (anthropometric proportions of the head and face, 1994).

Are high cheekbones actually more attractive?+

Cheekbone prominence reliably contributes to attractiveness ratings in faces of both sexes in the published literature, but the effect is non-linear. Perrett, Lee, Penton-Voak, et al. (1998) showed that moderate cheekbone prominence increased ratings while extreme prominence reduced them. Cunningham (1986) found cheekbone width and prominence ranked among the top correlates of female facial attractiveness ratings. The effect is also sex-dimorphic: high projection with narrow lower face cues femininity; high projection with wide lower face cues masculinity. Pure width or pure projection alone produces weaker effects than the combined signal.

Can I change my cheekbones without surgery?+

Bone structure does not change in adults. Three soft-tissue levers, however, change visible cheekbone prominence meaningfully. First, submalar fat loss (body composition reduction) deepens the hollow below the malar apex, increasing visible projection. Second, masseter and buccal volume reduction (dietary, postural, in some cases botulinum toxin to the masseter) narrows the lower face and increases the contrast against the upper midface. Third, contouring and grooming (subtle bronzer under the cheekbone, hairline shaping that frames the upper face) can move perceived prominence by a measurable amount in photo contexts. None of the three changes bone; all three change the visible signal that the bone produces.

What is a good cheekbone score?+

On the 17-metric scan, cheekbone prominence is reported as a composite of bizygomatic width index (cheekbone width over face width), malar projection index, and submalar contrast index. A composite at the 60th to 80th percentile of the published population norms is the band most consistently read as "high cheekbones" by raters. Below the 40th percentile reads as flat or full midface. Above the 90th percentile starts to read as gaunt or skeletal depending on the surrounding tissue distribution. The free scan returns your composite percentile and the three sub-scores so you know which axis to target.

Why does my photo show different cheekbones at different angles?+

Photo angle changes apparent malar projection by 20 to 40 percent depending on the head turn. A pure frontal photo flattens the projection axis; a three-quarter turn at 30 to 45 degrees produces the highest visible projection because the malar apex catches forward light while the submalar region recedes into shadow. This is the standard "good angle" effect. The 17-metric scan normalizes for head pose using facial landmark geometry, so the returned projection score is comparable across photos taken at different angles. Casual mirror or selfie comparisons are not normalized and will mislead you about real change.

How does cheekbone score interact with my full face score?+

Cheekbone prominence is one of 17 metrics in the full scan. It correlates moderately with attractiveness ratings on its own but contributes more in combination with three other metrics: midface ratio, lower-third length, and submental projection. A high cheekbone score paired with a long lower third reads as horse-faced; a high cheekbone score paired with a short lower third reads as compressed. The full report shows the cheekbone score in the context of the other midface metrics so you know whether the cheekbone is the lead lever or whether one of its neighbors is the real bottleneck.

What does the $14.99 PDF unlock that the free score does not?+

The free score returns the cheekbone composite percentile and the three sub-scores (width, projection, hollow). The $14.99 PDF report adds a written-by-aesthetic-coach interpretation of your specific cheekbone configuration, a ranked list of the soft-tissue and contouring levers that move your specific weak axis, a 30-day action plan with weekly photo checkpoints, and the comparison against the 16 other metrics so you know whether to prioritize cheekbones or one of the other midface levers.

Three axes. Five pages. Written by an aesthetic coach.

Unlock the written cheekbone report.

$14.99 unlocks the full 17-metric PDF: cheekbone configuration analysis, ranked levers, 30-day action plan, and the comparison against the 16 other metrics.

Unlock The Report — $14.99 →or try the free score first

Score your cheekbones on all three axes

Free, instant, private. Width, projection, hollow depth — plus 14 more metrics.

17 metrics · Photos auto-deleted · Re-scan as often as you want

Score My Cheekbones — Free →Unlock Full Report — $14.99

Improve your results

💪

Looksmaxxing Tips for Men

12 evidence-based improvements ranked by ROI

Read guide →🛒

Best Looksmaxxing Products

What actually moves the needle vs. what's hype

Read guide →✨

Glow-Up Guide

The fastest appearance improvements by category

Read guide →

Cheekbone test

The three axes the scan measures

What changes which axis

Bizygomatic width

Malar projection

Submalar hollow depth

Perceived width ratio

Perceived projection

Perceived cheekbone score

Why the three axes matter separately

Honest limits

Cheekbone test FAQ

Unlock the written cheekbone report.

Score your cheekbones on all three axes

Related Tools

Try our other tools

Smile Analyzer

Golden Ratio Test

Face Score

Face Rating

Attractiveness Test

Symmetry Test

Rank My Photos

Headshot Ranker

Looksmaxxing Test

Related on RealSmile

Face Rating

Face Rating Scale

Face Score

Face Shape Calculator

Glow-Up Guide

Glow-Up Roadmap