How to measure your canthal tilt: a simple self-test

To measure your canthal tilt: take a straight-on photo at eye level, keep a neutral face, then draw a line from the inner corner of your eye to the outer corner. Outer corner higher than inner = positive tilt; lower = negative; level = neutral. That's the entire test. The part the forums skip is bigger than the method: the number moves more with your camera and head angle than with your actual anatomy, which is exactly why you shouldn't read it as a verdict.

If you're here, you probably got handed a "negative tilt" call by an app, or you're about to run the canthal tilt test on yourself and you want to do it right. Fair enough — I'll give you the clean method. Then I'll show you why your own measurement is shakier than you think, and where the honest answer actually lives.

How do you measure canthal tilt at home?

You need one straight-on photo and a straight line. Shoot at eye level, look dead into the lens, hold a relaxed neutral face, then draw a line from the inner eye corner (the medial canthus) to the outer corner (the lateral canthus) and see which way it slopes. Up toward the temple is positive, down is negative, flat is neutral.

Here's the step-by-step, the way you'd do it to actually get a clean read:

1. Set the camera at eye level, dead front-on. Not above you (that fakes a downward, negative-looking slope), not below (fakes positive). Prop the phone, use a timer, and line your pupils up level with the lens.
2. Neutral face, eyes relaxed and open. No smile, no squint, no eyebrow raise. A grin lifts your cheeks and can swing the apparent angle by a few degrees on its own.
3. Good, flat, even light. Front light, nothing dramatic from the side. Shadows around the eye corner make the outer canthus hard to locate.
4. Find your two corners. The medial canthus is the inner corner near the tear duct; the lateral canthus is the outer corner. These are the two points the whole measurement rests on.
5. Draw the line and compare to level. Connect the two corners. Hold it against a true horizontal — the bottom edge of the frame, or a horizontal line you drop across both pupils. The direction of the slope is your tilt.

That's the honest version of every "measure your canthal tilt" tutorial. A photo, two dots, one line.

The manual method vs a measuring app

Two ways people do this, and they fail in the same place. Manually, you screenshot your photo, open any markup tool, and draw the corner-to-corner line yourself against a horizontal. An app claims to auto-detect your eye corners and spit out a number in degrees. The manual method at least makes the guesswork visible — you can see you're eyeballing where the outer corner sits. The app hides that same guesswork behind a confident decimal.

Neither is more accurate than the photo underneath it. If the photo has head tilt or lens curve baked in — and phone selfies always do — both methods faithfully measure the distortion and hand it back to you as your "tilt."

What is a normal canthal tilt in degrees?

There isn't a clean single number, and that's not a dodge. Most human eyes read somewhere between level and mildly upward — a few degrees of positive tilt is common — but the spread is wide and plenty of good-looking faces sit at neutral or slightly negative. No whitelist-grade research pins a specific degree count to how attractive you actually read.

So chasing a target number is a bit of a trap. You'll see forum figures thrown around like they're clinical thresholds, but they're mostly derived from measuring celebrity photos — which carry all the same camera and angle distortions you're about to introduce into your own selfie. A "target degree" built on distorted inputs isn't a standard. It's a rumor with a decimal point.

What's genuinely useful to know is the direction and roughly how strong it is — which is what positive vs negative canthal tilt breaks down. Precision past that is false precision.

Key numbers

• People form a stable read of a face — attractive, trustworthy, dominant — in about 100 milliseconds, and longer looks barely move it (Willis & Todorov, 2006). Nobody runs a corner-to-corner angle measurement in that window.
• A large review pooling eleven meta-analyses found strong agreement — within and across cultures — on who's attractive, judged holistically rather than by scoring isolated sub-traits like eye-corner angle (Langlois et al., 2000).
• People pull accurate impressions from just a few silent seconds of expressive behavior — movement, gaze, ease (Ambady & Rosenthal, 1992). A frozen tilt angle captures none of it.
• Across 37 cultures and roughly 10,000 people, a dependable character ranked among the top traits both sexes wanted in a long-term partner, and women weighted good financial prospects more heavily than men did — nowhere on that list is facial micro-geometry (Buss, 1989).

Why does my canthal tilt look different in every photo?

Because the measurement reads the photo, not your face — and the photo changes constantly. Head tilt, camera height, lens distortion, and a slight smile each shift the apparent angle by more than your actual anatomy varies from person to person. Run the test on three selfies from the same day and you can get three different answers.

This is the part that should reframe the whole exercise. The tilt you measure is a product of at least four things, and only one of them is your eyes:

Source of the number	How much it moves the result	Is it actually you?
Head pitch (chin up/down)	Large — a few degrees of nod flips the read	No, it's your pose
Camera height vs eye level	Large — shooting from above or below skews it	No, it's your setup
Lens distortion (phone front cameras)	Moderate — wide lenses curve the frame	No, it's your hardware
Smile / squint / brow	Moderate — cheeks lift, corners shift	No, it's your expression
Your resting eye anatomy	Small — a handful of degrees, fixed	Yes — but it's the quietest input

Look at that stack. The thing you're trying to measure is the smallest contributor to the number you get. Everything louder is an artifact of how you took the picture. A canthal tilt test done on one casual selfie is measuring your phone's lens and your chin position at least as much as your eyes.

Selfie front cameras are especially bad for this — they use short focal lengths that bend the geometry of a face held close, which is why the same person can look different in a selfie versus a photo shot from across the room. If you want to see how much your gear and framing distort your face generally, what do I actually look like goes into it.

Does the measurement even matter once you have it?

Barely — and this is where I'll stop being neutral, because the looksmaxxing framing here is just wrong. Even a perfectly measured canthal tilt is one faint static cue on a face that people experience in motion. It's near-invisible once you're talking, and no one you meet is running a protractor on your eye corner across a table.

The forums treat tilt like a load-bearing pillar of a "good face." The evidence doesn't. People judge faces as a whole, fast — Willis and Todorov (2006) found a 100-millisecond glimpse produces the same read as unlimited time, and no step in that process isolates your lateral canthus to grade it. Canthal tilt is also one of the worst candidates for a "key feature" precisely because it barely survives a smile: lift your cheeks and the angle changes. A real person meets the moving version of your eyes; the measurement only ever meets a frozen frame.

So the honest first impression is never "what's my tilt in degrees." It's a combination read — the whole face, plus body, plus how you're dressed, plus posture, plus the vibe you give off in the first second. Tilt is a rounding error inside that. The what is canthal tilt explainer makes the same case from the anatomy side; this piece is just showing you that even the measurement can't be trusted enough to justify the anxiety.

What should I do with the result instead?

Note the direction, then put the protractor down. If you measured a rough tilt and you're curious, fine — knowing you read as neutral or mildly positive or slightly negative is a harmless fact about your anatomy. What's not harmless is treating one number off one selfie as a diagnosis and reorganizing your self-image around it.

Here's where your attention actually pays off, ranked by how much it moves a real first impression:

• A relaxed, present expression. Does more for your eye region than any degree of tilt. Warm, un-tense eyes that hold a beat of contact and crease when you smile — that's most of the read, and it's fully in your control.
• Light and camera angle. The same variables that wreck your tilt measurement also decide whether a photo flatters you. Eye-level, soft front light, shot from a normal distance.
• Grooming, posture, body composition. Slow, controllable, real. These move first impressions in a way eye-corner geometry never will.

None of those require touching your eyes, and all of them beat the metric you came here to measure. If a tilt call or an app score has been living in your head rent-free, the most useful move is to point the question at something you can act on — which is exactly what the free test does. It reads how you land in that first second from a real-world perspective, instead of grading one line across one eye.

The bottom line

Measuring your canthal tilt is genuinely simple: eye-level neutral photo, draw a line from the inner corner to the outer corner, read the slope — positive up, negative down, neutral flat. That's the method, and now you have it. The catch is that your own measurement is dominated by head pitch, camera height, lens distortion, and expression, so the number describes your photo far more than your face.

And even a clean reading matters less than the forums swear it does. There's no whitelist-grade evidence that degrees of eye tilt move how women see you, the angle barely survives a smile, and people read your whole moving face in about 100 milliseconds (Willis & Todorov, 2006; Ambady & Rosenthal, 1992). Measure it if you're curious, then drop it. A cue this faint and this easy to mismeasure was never the thing to fix. If you want a read you can actually use, take the honest test — it skips the protractor and tells you which controllable lever is worth the most.

Worth reading next: what is canthal tilt and positive vs negative canthal tilt.

---

Studies referenced: Willis, J., & Todorov, A. (2006). First impressions: Making up your mind after a 100-ms exposure to a face. Psychological Science, 17(7), 592–598. Langlois, J. H., et al. (2000). Maxims or myths of beauty? A meta-analytic and theoretical review. Psychological Bulletin, 126(3), 390–423. Ambady, N., & Rosenthal, R. (1992). Thin slices of expressive behavior as predictors of interpersonal consequences. Psychological Bulletin, 111(2), 256–274. Buss, D. M. (1989). Sex differences in human mate preferences. Behavioral and Brain Sciences, 12(1), 1–49.