CEPHRA

Upper Facial Third

The share of the full face height taken up by the forehead, from the hairline down to the brows.

Drawn on a sketched reference face. Solid: measured lines · dashed: helper lines.
Formula
Height from hairline to brows ÷ full face height × 100
Unit
Percent (%)
Photo
Front view

How it is measured

  1. Front view: find the hairline on the midline and the lowest point of the chin.
  2. At each brow’s inner end, mark halfway between its upper and lower edge; the brow level lies midway between both marks.
  3. Divide the hairline-to-brow height by the hairline-to-chin height and multiply by 100.

Reading the value

Lower
Hairline and brows lie close together relative to the full face height.
Higher
Hairline and brows lie far apart relative to the full face height.

Why it matters

Splitting face height into thirds is a long-standing way to describe vertical balance; the three thirds shown here add up to 100 %. Measured faces rarely divide into equal parts, and the forehead share depends heavily on the hairline, which moves with age and hair loss.

Limits

The hairline is read from the contrast between skin and hair: light hair gives a weaker edge, and a fringe or shaved head leaves the value empty. Raised brows shrink this share. The chin bottom shifts with light.

Use a level, directly front-facing photo with even light. Head rotation, expression, hair or shadows can hide or shift the points. A measurement describes a photo, not a diagnosis, and no single value is good or bad on its own.

Video guide

Reference example · the values and scores shown belong to the example face.

Research

  • Hairline-lowering forehead reduction: pooled complications and satisfactionAesthetic Plast Surg · 2026 · meta-analysis · Strong evidence

    The pooled complication rate was 5.9%. Scarring occurred in 5.5%, inflamed hair follicles in 25.1%, hair loss at the scar in under 1%, revision surgery in 1.3%, and temporary numbness in 97.3%. Pooled satisfaction was 99.9%.

  • Forehead reduction (hairline lowering): about 1.6 cm and few complicationsFacial Plast Surg Aesthet Med · 2022 · meta-analysis · Strong evidence

    Mean forehead reduction was 1.6 cm (95% CI 1.4-1.8). Reported complications included temporary or permanent hair loss, poor scarring, lasting numbness and blood collection, with a pooled rate of 1% or less.

  • How facial proportions differ between ethnic groupsPlast Reconstr Surg · 2011 · systematic review · Strong evidence

    Forehead height showed the largest differences between groups, and distance between the eyes and nose width also varied notably. No significant sex difference was found in the classical proportions.

  • Hairline lowering: how stable is the shorter forehead?Laryngoscope · 2026 · cohort study · Moderate evidence

    The upper facial third fell by about 3.5% at the midline and 3.1% at the sides and stayed stable past 12 months; complications occurred in 26%, mainly temporary hair loss (14%), and 10% needed further surgery.

  • Facial proportions of young Black South African womenPlast Reconstr Surg Glob Open · 2023 · cross-sectional study · Moderate evidence

    The upper third of the face tended to be smaller than the middle and lower thirds. Nose width exceeded the classical rule of one fifth of face width and was larger than in other groups, tip projection was greater than in Congolese and African American samples, and upper and lower lip heights were close to 1:1 with the lips ahead of the E-line.

  • How the face changes with age in Korean women: 3D scansAnn Dermatol · 2021 · cross-sectional study · Moderate evidence

    Compared with the youngest group, the oldest group had a narrower eye opening, a longer upper eyelid area, a wider nose, a longer lower face, a longer and wider philtrum, a thinner lower red lip and a wider mouth. Nasofrontal, nasomental and mentolabial angles were larger and the nasofacial angle smaller.

11 studies in the research library are linked to this measure; the strongest are shown. These studies describe the feature or how it is measured. They do not show that one value is better than another.

Procedures that change it

Each example uses the app’s starting strength on one reference face; on another face the same procedure moves this value by another amount.

Measure your upper facial third – freeMeasured in your browser. Your scans are saved on your device.