Canthal tilt: the complete guide
Canthal tilt is the angle between the inner corner of the eye (the medial canthus) and the outer corner (the lateral canthus), measured against the horizontal. When the outer corner sits higher, the tilt is positive; when the corners are level, it is neutral; when the outer corner sits lower, it is negative.
What canthal tilt is
Each eye has two corners. The inner corner, nearer the nose, is the medial canthus; the outer corner, toward the temple, is the lateral canthus. Draw an imaginary line connecting those two points and compare its angle to a level horizontal line. That angle is the canthal tilt. It is a single, well-defined geometric measurement, which is what makes it one of the few "eye aesthetics" terms that maps cleanly onto something you can actually quantify.
The tilt is set mostly by where the lateral canthus attaches to the bony orbit and by the supporting tendons and tissues that hold the eyelids in place. Because the inner corner is anatomically more anchored than the outer corner, most of the visible variation — and most of the change over a lifetime — happens at the outer corner.
Positive, neutral, and negative tilt
The three categories describe the same measurement at different values. The approximate ranges below are hedged on purpose: published figures differ by method, landmark definition, age group, and population, so treat them as orientation rather than fixed cutoffs.
| Category | Outer corner vs. inner | Approximate typical angle | Common impression |
|---|---|---|---|
| Positive tilt | Higher | Roughly +4° to +8° (mild); higher is pronounced | Alert, lifted, "upturned" |
| Neutral tilt | Level | Around 0° to +3° | Even, balanced |
| Negative tilt | Lower | Below roughly 0° | Softer, "downturned," sometimes read as tired |
How to measure canthal tilt, step by step
If you want to estimate your own tilt from a photo, the method is straightforward, but the setup is where accuracy is won or lost.
- Get a level, front-facing photo. The camera should be at eye height and directly in front of you, with your head neither tilted nor rotated. Any head tilt adds or subtracts directly from the measured angle.
- Relax your face. Smiling, squinting, or raising the brows moves the soft tissue around the eye and changes the apparent corner positions. Use a neutral expression.
- Mark the medial canthus. Find the inner corner where the upper and lower lids meet near the nose.
- Mark the lateral canthus. Find the outer corner where the lids meet toward the temple.
- Draw the line and the horizontal. Connect the two canthi, then draw a true horizontal reference through the inner corner.
- Measure the angle. The angle between those two lines is the canthal tilt. Positive if the outer corner is above the horizontal, negative if below.
- Repeat for both eyes. The two eyes are rarely identical; measuring each separately is more honest than assuming symmetry.
The single biggest error source is pose. A five-degree head tilt turns into a five-degree error in the reading, which is enough to flip a "neutral" eye into "positive" or vice versa. This is exactly the limitation that depth-aware measurement is designed to remove.
Why canthal tilt trends online
Canthal tilt became a fixation in appearance-focused communities for a simple reason: it's one of the rare facial features that sounds objective. You can put a number on it, which gives discussions the feel of precision. It also anchors bigger aesthetic labels — for instance, a positive tilt is one of the four features bundled into the "hunter eyes" look — so it gets pulled into every conversation about eye shape.
The measurability is a double-edged sword. Because the number feels exact, people over-trust casual measurements taken from tilted selfies, and they chase a specific decimal target as if a couple of degrees were transformative. In practice, most of the variation people obsess over is within the noise of how the photo was taken.
What the research does and doesn't say
The honest summary is that canthal tilt matters a little, in context, and the popular framing overstates it. Here is a fair reading of the evidence.
- What research supports: a mildly positive tilt is the statistical norm in adults, and in some perception studies faces with a slight positive tilt receive modestly higher ratings than the same faces edited to a downturned tilt. Surgeons also use canthal position as one landmark in reconstructive and cosmetic planning.
- What research does not support: the idea that a specific tilt value is decisive for attractiveness, or that it can be ranked in isolation. Effects are small, they interact with every other feature, and they depend heavily on the rest of the face and on the observer.
- What is genuinely uncertain: precise population averages, because studies use different landmarks and imaging methods, producing a spread of reported values rather than one agreed figure.
The through-line, consistent with our writing on AI face analysis, is that no single measurement carries a whole face. Canthal tilt is one input among many, and treating it as a verdict misreads what the science actually shows.
Canthal tilt versus the whole eye
It helps to separate canthal tilt from the features people often confuse it with. Tilt is strictly the angle between the two eye corners. It is not the same as the eye's overall shape (almond, round, hooded), not the same as how much sclera shows above or below the iris, and not the same as brow position, though all of these influence the impression an eye gives. Two people can share an identical canthal tilt yet look completely different because their eyelid coverage, eye spacing, and brow height differ.
This matters because online discussions frequently attribute an eye's whole character to tilt alone, when tilt is only one contributor. If you measure your tilt and it doesn't match the "look" you expected, the explanation is usually that the other periorbital features — not the corner angle — are doing most of the visual work. Isolating tilt as a single clean measurement is exactly what stops it from being blamed for, or credited with, effects it doesn't produce.
How canthal tilt changes with age
Tilt is not fixed for life. The outer corner of the eye and its supporting tendon gradually descend with age as tissues lose elasticity and volume shifts. A tilt that is clearly positive in the teens and twenties often flattens toward neutral in middle age and can become slightly negative later. This is a normal, universal part of periorbital aging rather than a defect. It's also one reason comparing your eyes to a much younger celebrity's is misleading: you may be comparing across decades of predictable soft-tissue change, not across bone structure.
Because both age and photography move the apparent tilt around, a person's "true" resting tilt is best understood as a range measured under controlled conditions, not a single lifelong number.
How AI measures canthal tilt
Refrakt's eye analysis measures canthal tilt from a 3D scan rather than a flat selfie, which removes the largest sources of error:
- Pose correction. The scan establishes the true horizontal of the head in three dimensions, so head tilt and camera angle no longer contaminate the angle.
- Consistent landmarks. The medial and lateral canthi are located the same way every time, removing the guesswork of hand-marking corners on pixels.
- Left and right, separately. Each eye is measured on its own and compared, so natural asymmetry is reported rather than hidden.
- A signed angle, in context. The result is a positive or negative angle in degrees, presented alongside the other eye measurements and the wider face — never as a standalone score to pass or fail.
Measuring it properly is the whole value: it turns a number that swings wildly from selfie to selfie into a stable, honest figure you can actually interpret. For how canthal tilt sits within a full facial read, see our guides to AI face analysis and the broader eye-area aesthetics that the measurement feeds into.
Frequently asked questions
What is a normal canthal tilt?
Most adults have a mildly positive tilt, with the outer corner a few degrees higher than the inner. Reported averages often fall around 4 to 8 degrees, but the figure varies by method, age, and population, so treat it as an approximation.
Is a positive canthal tilt more attractive?
A mildly positive tilt is common and often rated favorably, but the effect is modest and context-dependent. Attractiveness rests on many correlated features, not one eye angle, and extreme tilt in either direction can look unnatural.
Does canthal tilt change with age?
Yes. The outer corner and its supporting tissue descend gradually over the decades, so a tilt that is positive in youth often flattens or turns slightly negative later. It's a normal part of periorbital aging.
How is canthal tilt measured accurately?
By marking the medial and lateral canthus, connecting them, and measuring the angle against a true horizontal. Accuracy depends on a level, front-facing pose, which is why depth-aware 3D measurement beats a casual photo.
Measure your canthal tilt properly
Refrakt reads canthal tilt from a pose-corrected 3D scan — a stable number instead of a selfie that changes every time you tilt your head.
Request early access