KBS Test: Korean Beauty Standards Test (2026)
Korean beauty standards are widely discussed, including in aesthetic-medicine writing — the V-shaped jaw, the small face, the high-bridge nose, the low-protrusion midface. This Korean beauty standards test runs the same scan as the main test: the same 17 measurements and the same Face Score. What it adds is a reading guide — which of those measurements relate to the traits Korean beauty discussions cite, which traits the scan cannot measure, and what moves each reading in a photo. It is descriptive, not prescriptive, and it is not a separate Korean score.
The 17 metrics — what matters most with Korean beauty standards in mind
Every attractiveness-test variant on RealSmile runs the same scan: 17 landmark-based measurements — symmetry, FWHR, facial thirds, midface ratio, canthal tilt, eye shape, brow arch, nose, lip, philtrum and chin proportions, jaw taper and more — plus your Face Score percentile. Nothing is re-weighted for this page. The table below is a reading guide: the measurements worth a closer look with Korean beauty standards in mind, and what moves each reading in a photo.
| Metric | What to know |
|---|---|
| Jaw Taper (V-Shape) | Jaw width against cheekbone width — the closest of the 17 measurements to the V-shaped jaw ideal. A lower ratio means a more tapered lower face; the scan's female reference range is 0.62–0.75 and the male range 0.72–0.85. Face the camera square-on at eye level. |
| Small face (소얼굴) — not measured | The 'small face' ideal compares the face with the whole head, and the scan has no landmark on the skull outline or hairline, so it cannot measure it. The nearest reading is FWHR (face width over upper-face height), which describes shape, not size. |
| Midface Ratio | Eye line to mouth, over brow to chin — a vertical proportion. The scan treats a more compact midface (0.42–0.48) as the youthful end of the range, the same direction a 'short midface' preference points. Because it is all-vertical, camera pitch shifts it: keep the lens at eye level. |
| Eye Shape (Hunter Eye Index) | Eye width over height. Narrower 'hunter' eyes give a higher index and a higher score; larger, rounder eyes a lower one. If the look you are comparing against favours large, round eyes, read a lower score here as rounder eyes, not a flaw. It moves with how open your eyes are, so keep a relaxed gaze. |
| Skin ('glass skin', 유리 피부) — not measured | None of the 17 measurements reads skin; all of them are landmark geometry. The Face Score is a learned read of the whole face in the photo, not of geometry alone, and nothing in the scan separates out what skin adds to it. Light still matters for the geometry: soft, even light helps the landmarks land accurately. |
Prefer to compare yourself with friends rather than with a benchmark? The mog rating scores everyone in one group photo.
Premium Dating Photo Audit · Delivered in 1–2 minutes
Want a written read of your photos?
The $19.99 Premium AI Dating Photo Audit scores each of your photos with the same scan, adds your validated Face Score percentile, and writes a 5-page commentary on which photo to lead with and why.
✓ 1–2 min delivery · ✓ 17 metrics measured · ✓ Lead-photo verdict · ✓ Stripe secured
What each score band means with Korean beauty standards in mind
The scan returns your Face Score, a 1–99 percentile. The score is the same one the main test gives — here is how to read it in four bands with Korean beauty standards in mind.
| Band | What it typically reflects |
|---|---|
| Top band85–99 | A top-band Face Score means the model fitted to human ratings places the photo near the top. It is the same score the main test gives, not a Korean-specific rating, and the 17 measurements are not part of it. The free result adds your strongest measurement; the full report shows jaw taper, midface and eye shape when your photo supports them, and the guide above relates each to the traits Korean beauty discussions cite. |
| High band70–84 | A high-band score sits one band under the top, and near the upper edge of the band a second photo of the same face can move it above or below that line. Retake with the camera level with your eyes and compare the two. A retake changes the photo, not your structure. |
| Mid band50–69 | A mid-band score sits in the middle of the range. Before reading it as your face, rule out the photo: a camera below the chin or a tilted head can change what the scan reads. Take two more photos, on different days if you can, and compare them. |
| Low band1–49 | A low-band score is a read of one photo, not a verdict on your face: across all the photos one person takes on the same day, the gap between their highest and lowest Face Score is a median of 11 points. Take two more photos at eye level, on different days if you can, and compare them before reading anything into it. |
Korean attractiveness test — FAQ
Is this Korean attractiveness test based on actual Korean preferences?
The score is not — it is the same scan as the main test, with no Korean-specific weighting. What this page adds is a guide to the measurements that relate to traits Korean beauty discussions commonly cite (the V-shaped jaw, the midface, eye shape), and it says plainly which traits the scan cannot measure (face size relative to the head, skin). The guide is descriptive (how the photo reads) rather than prescriptive (what it should look like).
What does a KBS percentage test actually measure?
Whatever traits and weights that site chose. No published, validated scale turns a face into a percentage match with Korean beauty standards, so two KBS percentage tests can give the same photo different numbers. This test does not print a KBS percentage. The number it gives is your Face Score, a 1 to 99 percentile from a model validated against 1,566 human-rated faces, and it is the same score the main test gives. The free result shows it with your strongest measurement; the full report adds the jaw taper, midface and eye-shape readings this page relates to Korean beauty standards, when your photo supports them.
Should I use this score to decide on cosmetic procedures?
No. The Korean attractiveness test is a photo-reading tool, not a cosmetic-medicine recommendation engine. We do not advise procedures of any kind, and any procedure decision should involve a licensed surgeon and an in-person consultation — never a single photo score. The score commentary is informational only.
Why does the score change so much between photos of the same person?
Because each reading comes from a single photo. Camera angle, head tilt, and lens distance all change what the scan reads. Across all the photos one person takes on the same day, the gap between their highest and lowest Face Score is a median of 11 points, and for a third of people it is 20 points or more. A gap that size is the photo, not your face — compare several photos before you believe any one of them.
Does this work the same for non-Korean faces?
Yes — the test runs on any face, and every face gets the same scan and the same score: 17 landmark-derived geometric ratios that work on any human face, plus the Face Score. Nothing is scored against a Korean-specific framework. This page only changes which measurements the guide explains and how it relates them to Korean beauty standards.
Are the references actually peer-reviewed?
There is no separate Korean reference set to review: the scan uses the same reference ranges on every page, and its research citations are listed on the RealSmile research citations page. The Korean beauty traits named here (the V-shaped jaw, the small face, the compact midface) are widely described beauty preferences; this page uses them to choose which measurements to explain, not to score you.
Related tests & tools
Premium Dating Photo Audit · Delivered in 1–2 minutes
See which of your photos reads best.
The free scan gives you your Face Score and your strongest measurement. The audit ranks your photos, picks your lead, flags the ones to cut, and writes a 5-page plan built on the geometry, not vibes.
✓ 1–2 min delivery · ✓ 17 metrics measured · ✓ Lead-photo verdict · ✓ Stripe secured
Other attractiveness-test variants