TL;DR — "Waterproof" and "smudge-proof" are unregulated words on makeup: the FDA keeps no list of approved cosmetic claims, and the one time it ruled on "waterproof" (sunscreen, 2011) it banned the word and replaced it with a tested "water resistant (40 or 80 minutes)" claim. Credible wear testing measures four stresses — friction, heat, moisture, and oil — and a claim that names no method, conditions, or threshold should be read as marketing, not measurement.
Pick two eyeliners off the same shelf, both labeled "smudge-proof," and try to find out which test either one passed. You can't — no U.S. authority defines "waterproof" or "smudge-proof" for makeup, so every wear promise on a liner or mascara label rests on tests the brand chose, ran, and graded itself. The useful skill here is not blanket distrust; it is knowing what a credible wear test actually measures — friction, heat, moisture, and oil — and checking whether the label tells you any of it.
A claim with no list behind it
Cosmetics reach the American market without pre-approval of their labels. The FDA states this plainly: the law does not require cosmetic labeling to have FDA approval before products go on the market, and the agency keeps no list of approved or accepted claims for cosmetics. The only general limits are that labeling must be truthful and not misleading, and that a product claiming to treat disease or change the body's structure or function becomes a drug in the eyes of the law.
"Waterproof," "smudge-proof," "budge-proof," "24-hour," "transfer-resistant" — all of these live in that unrefereed space, alongside older residents like "hypoallergenic." That does not mean the claims are false. Plenty of manufacturers run serious internal panels. It means there is no shared method, no shared pass mark, and no mandated wording. Two eyeliners can both say "smudge-proof" after clearing two entirely different bars — or, legally, after clearing none.
The one time "waterproof" met a referee
Sunscreen shows what happens when a regulator takes the same vocabulary seriously. In its June 2011 labeling rule, the FDA decided that manufacturers cannot label sunscreens "waterproof" or "sweatproof," because those claims overstate effectiveness; a product may only say "water resistant" if it passes the required SPF test procedure, and the front label must state whether it stays effective for 40 or 80 minutes of swimming or sweating.
Read that rule as a template and you see the anatomy of a regulated claim: a named test method, a fixed duration, and exact mandated wording. The absolute word "waterproof" did not survive contact with a stopwatch.
A second precedent shows the other direction. In 2016 the FDA asked makers of consumer antibacterial washes to back the word "antibacterial" with evidence of a benefit over plain soap and water; when the evidence did not arrive, the active ingredients behind the claim were removed from the market — a rule our tool-cleaning piece revisits from the other side. The pattern holds across categories: regulated words carry numbers and methods; unregulated words carry adjectives.
The four stresses the eye area actually applies
I spend a large part of my working life reading test standards and materials datasheets, and the design habit this article argues for is simple: define the environment before you trust the promise. The environment around the eye is a demanding test chamber, and it attacks makeup along four distinct axes.
Figure 1 — Four stresses, one word
Oil. The eyelid margin houses the meibomian glands, and their oily secretion is not incidental — the tear film's outer lipid layer is derived predominantly from meibum and is spread across the eye with every blink, according to the TFOS DEWS II report on tear film physiology. Liner near the lash line sits directly in the path of that oil, and the pilot data suggest migration takes minutes, not hours — the mechanism, and the three-subject caveat attached to that number, are covered in why eyeliner smudges.
Moisture. Tears, humidity, sweat, rain. This is the one axis the word "waterproof" gestures at — and, notably, the only one.
Heat. Skin-surface warmth, plus whatever the day adds. Heat softens waxes and film-formers and accelerates the oil axis.
Friction. Blinking is a repeating mechanical cycle, and hands, glasses, and pillows add irregular ones. Friction also dominates removal: a clinical study of 35 women with mild atopic dermatitis found that switching to a cleanser that required less rubbing force measurably improved dryness, irritation, and water loss within four weeks — evidence that the mechanical term in the equation is large enough to show up on skin.
A credible wear test exposes a formula to these four stresses separately and in combination. A marketing word collapses them into one.
A standard on paper, next to a slogan
The difference between a tested claim and a slogan is visible in the paperwork. Compare the wear vocabulary of makeup with claims that do have referees:
| Claim on a label | Referee | Named method | Pass threshold |
|---|---|---|---|
| "Water resistant" (sunscreen, US) | FDA 2011 final rule | Required SPF water-immersion procedure | Effective after 40 or 80 min |
| "Broad Spectrum" (sunscreen, US) | FDA 2011 final rule | Required test of UVA protection relative to UVB | Pass to carry the label |
| "Disinfects" (soft surfaces, US) | EPA registration guidance | Prescribed laboratory protocol | ≥4.0-log reduction in ≤10 min |
| "Waterproof" / "smudge-proof" (makeup, US) | None | None required | None |
Two things about that table repay a closer look. The EPA row exists because most registered disinfectants are approved only for hard, non-porous surfaces, and a soft-surface claim demands a minimum mean 4.0-log reduction in ten minutes or less — a claim defined entirely by a number, a time, and a surface class. And the two sunscreen rows come from a single 2011 rule: each word is tied to one named method and one threshold, printed on the front of the label. Any wear claim that omits its method, conditions, and threshold has omitted the result.
What independent tests find when they look
Third-party researchers have put "waterproof" formulas on the bench, and their findings cut both ways. A 2015 laboratory study coated silicone hydrogel contact lenses with cosmetics and measured the optics: every mascara tested reduced optical performance, waterproof formulations degraded it most, and cleaning restored the lenses only partially. That is a bench result, not a wearer trial — but it is an independent instrument reading, which no adjective on a label is.
The flip side of resistance is removal. A cross-sectional survey of 128 mascara users — self-reported, so another result to hold loosely — found the highest rate of eyelash fall, 27%, in the group that removed waterproof mascara with plain water, the one solvent a water-resistant film is built to defeat.
Put together, the independent evidence says the resistance is often real — real enough that the same film which shrugs off tears also shrugs off gentle removal. The label advertises the first half of that sentence and stays silent on the second.
The conditional bottom line
The honest reading rule is conditional. If a wear claim names its method, its conditions, and its threshold — as sunscreen labels now must — treat it as tested and let the numbers set your expectations. If it offers only the adjective, treat it as marketing shorthand for "more resistant than our regular formula, under conditions we chose," and assume the standard advice still applies: wear varies with oil, heat, moisture, and friction, and a film built to resist water needs something other than water to take it off.
Figure 2 — Reading a wear claim
That conclusion also tells you what to read next. Once you know the four stresses, the practical question stops being about adjectives and becomes which repair actually works when a liner loses to its environment mid-afternoon — the common fixing methods can be ranked precisely by how they handle oil and friction. And one step further down the same road sits a colder comparison: for close-up repairs at the lash line, the criteria that separate a dedicated precision tool from a bare fingertip are the same ones a good test protocol would use — control, hygiene, and how much force the surrounding skin absorbs.
