TL;DR — The common explanation for smudged eyeliner — bad formula, no primer, no setting powder — describes the smallest part of the problem. The eyelid margin is an oil-producing structure, and the film that oil forms is spread across the eye every time you blink, which puts anything drawn near the lash line directly in the path of a moving liquid. How much your liner actually travels depends on where you place it, what the product is made of, and your own lid physiology — and the published evidence on all three is thinner and more contradictory than the confident advice suggests.
There is a standard explanation for why eyeliner ends up somewhere under your eye by early afternoon, and it goes roughly like this: the pencil was too soft, you skipped primer, you forgot to set it with powder. Buy better, layer more, and the problem goes away.
I want to make the case that this explanation is not so much wrong as badly proportioned. It describes the last ten percent of the problem and ignores the first ninety. The first ninety percent is that you drew a line along the one part of your face whose job is to produce oil, and then you blinked all day.
I have not interviewed anyone for this piece. Everything below comes from published work — journal papers, a professional-society workshop report, and patient-education material from ophthalmologists and dermatologists — and I have tried to be honest about how strong or weak each piece of it is, because in this particular corner of the literature the confidence of the advice tends to outrun the size of the studies.
Why does eyeliner smudge even when the formula is good?
Because the lid margin is not a neutral surface. Along the rim of both eyelids, behind the lash roots, sits a row of sebaceous glands — the meibomian glands — and they secrete an oily substance called meibum. That secretion is the source of the outermost layer of the tear film, and the mechanism that distributes it is the blink itself. The Tear Film & Ocular Surface Society's DEWS II report on the tear film puts it plainly: the lipid layer is derived predominantly from meibum and spreads upward over the watery layer with blinking.
Read that sentence again with makeup in mind. There is a thin, mobile, oily film being pumped out along the lash line and swept across the surface of your eye a few times every minute, for as long as you are awake. A line of pigment placed at or near that margin is not sitting on dry skin. It is sitting in the runway of a liquid that is constantly being re-spread.
This is why "set it with powder" is such an appealing answer and such an incomplete one. Powder addresses sebum on the eyelid skin above the lash line. It does very little about the film emerging from the gland orifices below it, and nothing at all about the mechanical action of the lid sliding over the globe several times a minute.
Does eyeliner really migrate, or is that just marketing?
It migrates — but the evidence is smaller than you would guess from how often it is repeated.
The study everyone is quoting, usually without saying so, is a slit-lamp experiment by Alison Ng, Katharine Evans, Rachel North and Christine Purslow, published in Eye & Contact Lens in 2015. They used a glitter-containing pencil liner as a visible tracer and watched where the particles went. When the liner was applied behind the lash line rather than on the skin outside it, the particles moved into the tear film more readily, with maximum contamination reached five to ten minutes after application.
Five to ten minutes is a striking number, and it is also a number produced by three subjects. This was explicitly a pilot study, the tracer was glitter rather than the pigments and waxes of an ordinary liner, and contamination had become negligible by two hours. It is real evidence that particulate matter crosses the lid margin quickly. It is not a population fact, and anyone who writes "studies show your eyeliner is in your tear film within five minutes" is quoting three people.
The most-cited synthesis of this area, a review by Michael Wang and Jennifer Craig in Clinical Optometry, is careful in the same way. It concludes that wearers should be advised to avoid applying product to the inner eyelash line in order to minimize migration across the lid margin — sound advice, but a narrative review resting on a small underlying literature, and it says so.
Is the waterline the whole story?
This is where the tidy version of the advice starts to fall apart, and it is worth watching it fall apart rather than pretending it doesn't.
Figure 1 — Neither placement came away clean
The American Academy of Ophthalmology's consumer guidance, reviewed by Rebecca J. Taylor, MD, is unambiguous: always apply makeup outside the lash line, away from the eye, to avoid blocking the oil glands of the upper or lower eyelid. That is a professional body's position, and it lines up neatly with the migration data above.
But a 2026 study in the Journal of Clinical Medicine by Alkawally, Lin, van de Pol, Sasai, Nguyen and Paugh compared three groups — people who wore no eye makeup, people who applied only outside the lid margin, and people who applied directly on the margin — and found that neither placement came away clean, but not in the same way: gland secretion quality was significantly worse only among those applying on the margin, while significantly increased ocular surface staining appeared only among those applying outside it. Outside the lash line was not a clean escape. In the same study, tear breakup time was identical across all three groups at 4.6 seconds, and there was no difference in gland dropout or lipid layer thickness.
The groups were small (10, 18 and 21 people), so I would not overturn anything on the strength of it. But it is the most useful paper in this set precisely because it refuses to confirm the neat story. "Waterline bad, outside the lashes safe" is a reasonable rule of thumb. It is not a mechanism, and it is not a guarantee.
What about the product itself — does chemistry matter at all?
Yes, but probably in a narrower way than the marketing implies.
There is one intriguing in-vitro finding. Hunter, Bhola, Yappert, Borchman and Gerlach mixed eyeliner products with human meibum in a spectrometer and found that one makeup product increased the lipid order, or viscosity, in a way that could have adverse effects on tear film stability — raising the lipid phase transition temperature by 4.2 °C. That is a candidate mechanism for a product physically changing the behavior of the oil it lands in.
Two caveats, both large. No eyes were involved; this happened in glassware. And of the two products tested, only one did anything at all — which makes this a property of a formulation, not of eyeliner as a category.
The more mundane compositional point is about long-wear claims. The TFOS Lifestyle report on cosmetics notes that acrylates and polymethyl methacrylate are used as film-formers in mascara and eyeliner. Those polymers are what make a long-wear formula stay put. A liner that resists water resists tears too. That does not make it immune to the mechanics of the lid — it just changes what it takes to remove it later, which is its own subject and one we cover in the cluster's removal guide.
Why does eyelid movement matter so much?
Two structural facts compound the oil problem.
First, the skin you are drawing on is the thinnest on your face. In a cadaveric study of 39 facial subunits, Chopra and colleagues reported that eyelid skin is the thinnest in the face, with the upper medial eyelid measuring a dermal thickness of about 759 micrometres. Thin, highly mobile skin creases with every blink, and creasing is a transfer mechanism — the upper lid margin makes contact with the lower lid area thousands of times a day.
Second, this is not a rare exposure. In a survey of 1,360 women by Ng and colleagues, 83% reported using eye cosmetics regularly, at least three times a week. Whatever the interaction between makeup and the lid margin turns out to be, it is happening at scale and daily. (That sample was young and self-selected, so treat the percentage as a description of that group, not of women generally.)
Does any of this mean eyeliner is damaging my eyelids?
Here is where I want to be careful, because this is the point at which articles like this one usually overreach.
Figure 2 — Tear breakup time: two studies, two answers
Two cross-sectional studies have found associations worth knowing about. Prabhasawat and colleagues, writing in Cornea, compared regular eyeliner users with non-users and reported that tear breakup time was significantly lower in the eyeliner group — 3.0 seconds versus 5.8. And in a larger sample of 220 healthy women, Z. E. Ercan found that the no-makeup group had significantly less meibomian gland loss than the eyeliner-only group on meibography.
Now the counter-reading. Both are cross-sectional: they show association, not causation, and women who wear liner daily may differ from women who don't in a dozen other ways. In the Prabhasawat study the control group's breakup time was already below the usual normal cutoff. In the Ercan study, tear production measured by Schirmer testing did not differ between groups, and combining eyeliner with mascara was not worse than either alone — which is awkward for any simple dose-response story. And the 2026 study above found no group difference in gland dropout at all.
It is also worth separating two ideas that often get merged. Gland obstruction is a real and well-described process: when orifices become blocked you get meibum stasis, cystic dilation and eventual gland atrophy, as Chhadva, Goldhardt and Galor describe in Ophthalmology. But the obstruction they describe is driven by hyperkeratinization of the duct, not by cosmetics. That paper explains what happens once a gland is blocked. It does not establish that eyeliner blocks it.
The honest answer is that we do not fully know. There is a plausible mechanism, several small studies pointing the same direction, and at least one careful study that declines to confirm the tidiest version of it.
So what actually determines whether your liner travels?
The conditional answer — which is the only defensible one — is that it depends on four things at once:
Placement. The closer to the lid margin, the closer to the oil source and the faster the migration observed in the pilot data. This is the single most consistent finding in the set, and also the one resting on the fewest subjects.
Formulation. Film-forming polymers resist water. One tested product measurably changed the physical behavior of meibum in a dish; another didn't. Product-level differences appear to be real and are not predicted by price.
Your own physiology. Meibum output, blink completeness, lid laxity and tear film stability vary enormously between people, which is very likely why one person's liner sits still for eight hours and another's is gone by lunch on the same product.
Mechanics. Blinking, creasing, and the sheer amount of contact between two moving surfaces of the thinnest skin on the body.
Notice what is missing from that list: whether you set it with powder. Setting powder is not useless — it absorbs surface sebum on lid skin, and that is worth something. It is simply not addressing the main variable. If you want the practical version of all this, the main guide in this cluster walks through what to do when the line has already moved, and the five-step correction covers the fix itself.
When is this a question for a professional rather than a beauty problem?
If the issue is not cosmetic — if you have persistent redness, itching, flaking of the lid skin, unexplained lash loss, or discomfort that outlasts the makeup — that is a question for an eye doctor or a dermatologist, not for an article.
Eyelid dermatitis in particular is multi-causal and genuinely difficult to attribute by guesswork. An American Academy of Ophthalmology evidence review by Dagi Glass and colleagues identified fragrance mixes, Balsam of Peru, cinnamic alcohol, gold, methylisothiazolinone, neomycin, nickel and personal care products as allergens with at least 10% positivity across multiple patch-test studies. Those span jewelry, topical antibiotics and nail products — and eyelid reactions are frequently caused by transfer from the hands rather than by anything applied to the eye. Which is a good reminder that blaming the eyeliner is a hypothesis, not a diagnosis.
I set out to write a piece explaining why eyeliner smudges. What I found was a mechanism that is well established in ophthalmology and almost entirely absent from beauty advice, wrapped around an evidence base that is smaller and more contradictory than either side tends to admit. Both halves of that seemed worth saying.
