The average face, measured
Published anthropometric figures, with their sources and their sample sizes — and the reason they contradict each other, which is more useful than any single number.
The short answer. There is no one answer, and the reason matters. 'Face length' means at least three different measurements, and the figures for them differ by about three centimetres. Published means also vary by population and by whether the face was measured alive or as bone. Any site quoting a single average face length without saying which landmarks it used is quoting a number that cannot be checked.
The question this page answers: how long and how wide is an average face?
The problem with the question
Search for the average face length and you will be given a number, usually to one decimal place, usually with no source. The numbers disagree with each other by several centimetres, and the reason is not that anyone is lying. It is that face length names at least three different measurements, and almost nobody says which one they mean.
| Measurement | From | Typical adult |
|---|---|---|
| Morphological face height | Nasion — the bridge of the nose, between the eyes — down to the chin | 11–13 cm |
| Physiognomic face height | Trichion — the hairline — down to the chin | About 3 cm more |
| Landmark face length | A software reconstruction of the top of the forehead, down to the chin. What face analysis apps measure, including ours. | Between the two |
Anthropometry standardises on the first, because the bridge of the nose is a bony landmark that can be found reliably on anyone. The styling literature uses the second, because the hairline is what a person looking in a mirror can see. The two differ by about three centimetres, and a guide mixing figures from both — which many do — produces advice that cannot be followed.
The hairline is the weakest landmark in this whole subject, and the reason is simple: it moves. It recedes with age, it differs between individuals at the same age, and on many people it cannot be located to better than a centimetre. Our own tape-measure tool treats the gap between the hairline and the reconstructed forehead top as 29 mm with a standard deviation of 8 mm — an uncertainty wide enough to change a classification, which is why we publish it rather than hide it.
Two published studies, side by side
Rather than quote one number, here are two real studies with their methods attached. They were chosen because they disagree, and the disagreement is instructive.
| Study | Sample | Cheekbone width | Jaw width | Face height |
|---|---|---|---|---|
| Swedish adults, living87 men, 77 women | Men | 150.0 mm | 133.4 mm | 120.3 mm |
| Women | 145.8 mm | 125.2 mm | 112.3 mm | |
| Nigerian adults, dry skulls78 male, 22 female skulls | Men | 139.2 mm | not reported | 126.2 mm |
| Women | 124.8 mm | not reported | 120.0 mm |
The men's cheekbone widths differ by 10.8 mm between the two studies. That is larger than most of the distinctions people are trying to draw when they measure a face at all, and it has two causes rather than one.
- Population. Facial proportions vary systematically between populations, and every published "average" is the average of whoever was in the room. Neither of these studies is a sample of humanity, and neither claims to be.
- Method. One measured living faces with callipers over skin and muscle; the other measured dry skulls. Soft tissue adds several millimetres to every width. A figure from one cannot be compared with a figure from the other, however carefully each was taken.
This is the heart of it. The absence of a single average face length is not a gap in the literature waiting to be filled — it is what the literature says. An average exists only relative to a population and a protocol, and a number quoted without both is not a measurement.
The classification nobody online uses
Anthropometry has had a face shape classification for over a century, and it is not the one styling guides use. The morphological facial index is face height divided by cheekbone width, times 100, with five published bands.
| Index | Name | In plain words |
|---|---|---|
| up to 78.9 | Hypereuryprosopic | very broad face |
| 79.0–83.9 | Euryprosopic | broad face |
| 84.0–87.9 | Mesoprosopic | medium face |
| 88.0–92.9 | Leptoprosopic | long face |
| 93 and over | Hyperleptoprosopic | very long face |
Running the two studies above through it produces a result worth sitting with:
| Study | Index | Band | |
|---|---|---|---|
| Swedish adults, living | Men | 80.2 | Euryprosopic — broad |
| Swedish adults, living | Women | 77.0 | Hypereuryprosopic — very broad |
| Nigerian adults, dry skulls | Men | 90.7 | Leptoprosopic — long |
| Nigerian adults, dry skulls | Women | 96.2 | Hyperleptoprosopic — very long |
By the oldest and best-established face shape classification there is, the two studies put the average adult face in opposite categories. Neither is wrong. They measured different populations by different methods, and the index — which has no way of knowing that — reports the difference as though it were about the faces.
What to take from this. A face shape result is only as meaningful as the protocol behind it. That is why every tool on this site shows the ratios and the thresholds alongside the label: the label is a convention, the ratio is the measurement, and only one of them survives a change of method.
What our own tools measure, and what they call average
Our face shape tools measure the top of the reconstructed forehead to the chin, over the widest point across the face. That is a third definition again, and we state it rather than implying it matches a tape measure, because it does not.
| Ratio | Typical | What it means |
|---|---|---|
| Length ÷ width | 1.15 | A tape from the hairline would say about 1.5 on the same face. The landmark set measures a shorter length over a wider width, so the number is smaller. |
| Forehead ÷ cheekbones | 0.86 | The forehead is narrower than the cheekbones on nearly everyone. |
| Jaw ÷ cheekbones | 0.8 | Published anthropometric norms put bigonial over bizygomatic width at roughly 0.77 to 0.79, so our 0.8 is close agreement. A jaw at 80% of cheekbone width is the ordinary jaw, not a narrow one. |
That last row is the most useful thing on this page for anyone reading face shape guides. A great many of them describe a narrow jaw as anything under about 92% of cheekbone width — a threshold that fires on essentially every human being, and the reason so many quizzes return "heart" or "diamond" to almost everyone. Thresholds have to be set against what an ordinary face measures, and ours are, which is why we publish the ordinary figures.
How to measure your own, properly
- Decide which length you mean and keep it consistent. Hairline-to-chin is easier to find on yourself; nasion-to-chin is far more repeatable.
- Use the widest point across your cheekbones, not across your temples or your head. Find it by feel, with your fingertips, before you reach for a tape.
- Measure across the face, not around it. Wrapping a tape round your head measures an arc, which is several centimetres longer than the straight-line width every published figure uses.
- Measure the jaw at the gonial angles — the corners below and in front of your ears — not at the chin.
- Take each measurement three times. Hand-held tape on a face is good to about 3 mm, and the spread between your three attempts tells you how much to trust the result.
Our tape-measure tool does this arithmetic for you, carries the uncertainty through to the answer, and tells you when your measurements are too close to a threshold for the result to be meaningful — which, given tape accuracy, is more often than you might expect.
A note on what averages are for
An average face is a statistical construct and nothing more. It is useful for designing spectacle frames, sizing helmets, and planning reconstructive surgery, which is what these studies were done for. It is not a target, and a measurement away from it is not a finding about anybody.
It is also worth saying plainly that composite "average faces" — the blended images that circulate as the average face of a country — are averages of the photographs that went into them, which is a different thing from an average of a population. They look smooth and symmetrical because averaging removes asymmetry and texture, not because the average person is smooth and symmetrical.
Published figures here are quoted from the two studies linked below, with their sample sizes and methods stated. They describe specific groups rather than humanity, and the measurements on this page carry no health, aesthetic or diagnostic meaning. Nothing here scores or ranks anyone's face.
Sources
- Swedish adults, living — 87 men, 77 women. Employees of a Gothenburg engineering firm aged 25 to 49, measured on living faces with spreading callipers over the soft tissue.
- Nigerian adults, dry skulls — 78 male, 22 female skulls. Fully ossified adult skulls, measured on bone with no soft tissue at all — so every figure is smaller than the same person would have measured alive.
Questions people ask
- How long is the average face?
- There is no single answer, because "face length" means at least three different measurements. From the bridge of the nose to the chin — the measurement anthropometry standardises on — published means run from about 11 to 13 cm for adults. From the hairline to the chin, the measurement most online guides mean, figures run about 3 cm longer, but the hairline is not a fixed landmark so that measurement is far less reliable.
- How wide is the average face?
- Across the cheekbones, published means for adults run from about 12.5 to 15 cm depending on population and on whether the measurement was taken on living faces or on bone. The two studies compared on this page differ by 11 mm for men, which is larger than most of the differences people are trying to detect.
- What is a normal face width to height ratio?
- By the classical facial index — nasion to chin over cheekbone width — the medium band runs from 84 to 88. Below that is a broad face and above it a long one. Using the hairline-to-chin measurement instead gives a ratio over 1.0, usually quoted around 1.3 to 1.5, which is why the two conventions look like they disagree.
- Is my face too long or too wide?
- Neither is a thing a face can be. These categories describe proportion and carry no health or aesthetic meaning, and the published averages come from specific populations rather than from humanity. A measurement outside a quoted range means your face differs from that sample, which tells you about the sample as much as about you.
- Why do online face measurement guides all give different numbers?
- Because they are measuring different things and rarely say which. Face length can be hairline-to-chin, nasion-to-chin, or a software landmark's reconstruction of the top of the forehead; width can be the cheekbones, the temples or the whole head. Mixing figures taken by different definitions is the single commonest error on this subject.