Eye wrinkles: why doesn't eye cream change them?

Eye cream mostly changes how much water sits in the top layer of skin, and water changes how a line looks within minutes. A wrinkle that has settled in sits deeper than that — in the dermis, where elastin is interrupted and collagen has thinned. Two different layers, on two different clocks. That gap is why the jar can be working exactly as designed and the line can still be there.

You buy the eye cream, you use it every night, and six months later the same line is in the same place. It's a fair thing to be annoyed about. It's also worth knowing what you were asking the cream to do.

What's under an eye wrinkle that has settled in?

Not a dry patch. In ex vivo human skin — 157 biopsies from 46 people — the dermis directly beneath a wrinkle showed interrupted elastin deposits and pronounced collagen atrophy, with a thinner epidermis over the top (Contet-Audonneau et al., 1999). A settled line is a structural change in the dermis, and nothing applied to the surface refills a dermis.

That's the honest starting point. It isn't a reason to stop caring for the eye area — it's a reason to be specific about what each thing in your routine is for.

Why does the eye area look smoother right after you apply something?

Because some of what you see is a water reading. In a study measured on the eyelids of 20 volunteers, 30 minutes at 70% humidity followed by 30 minutes at 40% humidity was enough to significantly lower skin conductance and elasticity and significantly increase fine wrinkles (Tsukahara et al., 2007). Thirty minutes is far too short for collagen to change, which means the count of visible fine lines moves with water long before it moves with structure.

So the smoothing you notice an hour after applying is real. It is also reversible, and it resets every time the air around you changes. The same effect on the rest of the face is covered in why fine lines look worse when skin is dehydrated.

Which ingredients actually have wrinkle evidence?

Retinoids, by a distance. A meta-analysis of 8 randomized controlled trials in 1,361 people found tretinoin improved fine wrinkles with a mean difference of 0.412 (95% CI 0.233–0.590) against control (Huang et al., 2025).

One caveat matters here, and it is rarely stated. That evidence is measured on the face as a whole; no placebo-controlled effect size has been published for crow's feet on their own. A split-face periorbital trial exists (Kim et al., 2010, 24 completers), but it tested retinyl retinoate — a niche ester, not a mainstream retinoid. How retinoids act on collagen is in what "boosts collagen" actually means.

Are peptides as proven as they sound?

Thinner than the marketing suggests. A meta-analysis of 19 randomized trials in 1,341 people found a pooled wrinkle reduction of 0.27 (p = 0.04) — but only 2 of those 19 trials tested peptides applied to skin. Seventeen were oral, and the oral subgroup, at a mean difference of 1.5, largely drove the pooled result (Nukaly et al., 2026).

The single trial dedicated to crow's feet is smaller still: 21 women in three arms of seven, eight weeks, twice daily to the periorbital area. Grading scores fell, but corneometer, tewameter and cutometer readings all came back non-significant (Aruan et al., 2023). Seven people per arm cannot settle a question like this. It isn't evidence that peptides fail — it's evidence that nobody has tested them properly on the eye area yet. The comparison with retinol is in retinol vs. peptides for eye wrinkles.

What an eye cream can reasonably do

Quite a lot, once you stop asking it for the wrong thing. It can hold water so lines read softer today. It can reduce the friction of rubbing and dragging, which the thinnest skin on the face does not need. It can carry a retinoid at a strength thin skin tolerates, if you choose one. What it cannot do is put back collagen that isn't there, and a product isn't failing when it declines to perform a miracle.

Set the expectation to match the layer you're working on, and the routine gets simpler rather than longer.

One step today

• Judge an eye product over 8 to 12 weeks, in the same light and at the same time of day — not an hour after applying.
• If indoor air is dry, keep humidity around 40–60%; it changes how fine lines read on the same day.
• If you want the ingredient with the strongest wrinkle evidence, that's a retinoid — start at a low strength, two or three nights a week.
• Apply with a patting motion rather than dragging; eyelid skin is the thinnest on the face.