No eye cream reaches the muscle that makes crow's feet.

The outer eye takes more sun than any other facial skin and gets less sunscreen — most people stop short of the lashes. The Nambour trial's daily-sunscreen group showed no detectable increase in skin aging over 4.5 years, 24% less than discretionary users (Hughes 2013), and the thin periorbital dermis is where that collagen protection matters most. Sunglasses add two things: they block the UV that reaches the periorbital skin — modelled doses vary widely with frame style, and goggles or large lenses protect the skin around the eye far better than medium-sized ones (UV-dose study) — and they stop the squint, which is the muscle folding the skin against glare thousands of extra times a day. No trial has randomised sunglasses against crow's feet; the mechanism is the entire cause of the line.

Practically: a mineral or gentle chemical SPF 30–50 to the orbital rim every morning, large UV400 sunglasses whenever it is bright, and a hat. The dermatologist's habit is to put the sunscreen on before make-up and to keep sunglasses in the car and the bag, because the squint happens in the two minutes before you remember them.

Retinoids are the one topical that thickens the epidermis and rebuilds upper-dermal collagen, and the periorbital skin is where that shows first. The meta-analysis of eight randomised tretinoin trials in 1,361 patients found significant improvement in fine and coarse wrinkles over vehicle (meta-analysis); a hydrating retinoid eye cream improved the appearance of periorbital lines by 33% at 12 weeks in an open-label study (JDD, 2022); and an 8-week randomised comparison in 196 women with periorbital wrinkles found a niacinamide–peptide–retinyl propionate regimen comparable to 0.02% tretinoin with better tolerance (regimen trial). It treats crepe and fine static lines, not the dynamic fold, and it needs months.

The eye tolerates less than the cheek: a pea for both sides, on the orbital bone rather than the lid margin, two or three nights a week to start, buffered with a moisturiser, and never into the eye. Retinol or retinaldehyde at a stated strength is the gentler route; the wrinkles guide covers the ladder.

Three ingredients with controlled data for fine periorbital-type lines: niacinamide 5% reduced fine lines and wrinkles and improved elasticity in a 12-week double-blind split-face trial of 50 women (Bissett 2005); a 5% vitamin C cream reduced deep furrows and improved surface microrelief over six months in a double-blind trial (Humbert 2003); and a multi-weight hyaluronic-acid serum produced immediate, measurable reductions in fine and coarse lines by hydrating the surface (JAAD, 2016). Around the eye they soften crepe, add antioxidant protection under sunscreen, and make the skin drape better over the toxin result. None reaches the muscle, and none fills a static line.

Red and near-infrared light nudge fibroblasts to make collagen, and crow's feet are the one wrinkle for which a home mask has been tested properly: a multicentre, randomised, double-blind, sham-controlled trial of a home-use LED and infrared mask found significant improvement in crow's feet over the sham device over the study period (sham-controlled trial), consistent with the older clinic-dose trial of red light for photoaged skin (Wunsch 2014). The improvement was judged by blinded raters on photographs and appeared from week eight of use five times a week, and comes with eye protection built into the device. A reasonable needle-free add-on for crepe and fine lines; not a substitute for toxin on a dynamic fold. Our red-light guide covers doses.

Among 79 pairs of identical twins discordant for smoking, the smoker scored worse on upper-lid skin, lower-lid bags and malar bags — though not on crow's feet — with five years of smoking history visible in photographs (twin study). Side and stomach sleeping press the temple and outer eye into the pillow for a third of life, and surgeons attribute a characteristic vertical or diagonal crease near the eye to it (Aesthetic Surgery Journal) — plausible, thinly evidenced, cheap to test with a pillow. Squinting at a small or dim screen is the orbicularis doing the sun's job indoors; glasses that are the right prescription and a bigger font are prevention nobody sells. None of it reverses a line; all of it decides the slope.

Argireline (acetyl hexapeptide-3) relaxes muscle in a dish and is sold as topical toxin; in a small double-blind randomised trial on crow's feet (21 women), palmitoyl pentapeptide-4 looked better on photographs and self-assessment, but no instrument reading was significant and the two actives did not differ significantly (JCAD trial). A cyclised hexapeptide-9 outperformed retinol against aging signs in a manufacturer-run randomised, vehicle-controlled trial (2025 RCT), and a multi-peptide eye serum improved crow's-feet measures in a 2023 controlled study (2023 trial). The honest summary: measurable, small, manufacturer-funded improvements in fine lines and hydration, and no peptide has ever been tested against toxin for a dynamic line, because a molecule that cannot reach the muscle cannot stop it.

Microneedle patches carry hyaluronic acid into the upper skin on dissolving needles, and in a randomised split-face study of 34 women with mild to moderate crow's feet, eight weeks of twice-weekly patches improved wrinkles more than the same hyaluronic acid applied topically (split-face RCT); a pilot of a near-infrared photothermal microneedle patch improved under-eye roughness and crow's feet over four weeks (pilot trial). Silicone patches hydrate and splint a crease overnight, with no controlled data; adhesive pads have been tested, and in a placebo-controlled study of 39 people they improved crow's-feet wrinkles "in the first hour after use" (placebo-controlled study). Both act on the surface; neither reaches the muscle or rebuilds a dermis.

The 20-week facial-exercise pilot in 16 women found fuller cheeks and a slightly younger rated age, with no control group and no wrinkle measurement (Alam 2018); in a small controlled study of 18 people, seven weeks of daily exercises changed nothing a lay panel could see except the upper lip, and the authors wrote "It cannot be concluded that facial exercises are effective" (De Vos 2013). For crow's feet the logic runs the other way: the line exists because the orbicularis folds the skin repeatedly, and the treatment with the best evidence in aesthetic medicine works by stopping exactly that. Eye-squeezing exercises add repetitions. Harmless for the cheeks; counterproductive at the corner of the eye.

Crow's feet are the second cosmetic indication botulinum toxin earned, and the trials are the model for the field. OnabotulinumtoxinA was licensed for lateral canthal lines in 2013 at 24 units — three points a side (FDA label). In the phase 3 programme, investigator and patient responder rates for crow's feet alone were 54.9% and 45.8% at a month against 3.3% on placebo, rising to 59% and 48.5% when the frown lines were treated at the same time (phase 3); repeated cycles held their effect (repeat-treatment study); the median duration was about four months (duration study); and satisfaction with the whole upper face stayed high for up to six months (2022 study). The early dose-ranging trial showed benefit at every dose from 3 to 18 units a side, with 12 units recommended (dose-response study).

Every newer toxin has repeated the result: prabotulinumtoxinA in a phase IV placebo-controlled trial (phase IV), the liquid relabotulinumtoxinA with 51.8% composite two-grade responders against 1.4% (READY-2), and daxibotulinumtoxinA with a median duration of 24 weeks at one dose in phase 2 (phase 2a); a 2026 GRADE-assessed meta-analysis covers incobotulinumtoxinA across the upper face (meta-analysis). Effect starts at day three to five, peaks at two weeks, and a lighter dose keeps the smile. The differences between brands are smaller than the differences between injectors.

Once a crow's foot is static, the muscle is half the problem and the floor of the line is the other half. In a randomised, controlled, double-blind split-face study of 25 women with moderate to severe periorbital lines, both sides received toxin and one side a microdroplet hyaluronic gel (Juvéderm Volite/Skinvive): the combined side did better on the Merz scale and held its result longer over six months (split-face RCT). An 18-month prospective, randomised, maker-funded, single-blind study of a superficial fine-line hyaluronic filler in 196 subjects found significant crow's-feet improvement from three weeks that remained significant at 18 months after a single session (18-month study); a non-cross-linked hyaluronic gel improved lateral canthal lines in an open multicentre study (open study); and the combination approach for the whole periorbital and temple region is described in a remodelling series (series). A randomised trial of toxin with non-cross-linked collagen reached a similar place (collagen RCT).

The gel goes in after the toxin has settled, very superficially, in tiny amounts, along the line and not the fat pad — and only soft, low-water products, because half-millimetre skin turns blue over anything stiffer. The filler guide covers the boosters and the risks.

The outer orbicularis oculi is a brow depressor; weaken it and the frontalis pulls the lateral brow up unopposed. Injecting 1–2 units just below the outer brow, above the bony rim, lifted the lateral brow by 0.6–2.1 mm in a randomised trial of 15 patients (30 brows) per group, with 97% satisfied (brow-height RCT), though how much any face lifts is unpredictable (StatPearls; anatomical study; review). One small randomised trial and the anatomy support it, which is why the botox guide grades it moderate; the effect is real in the right face and invisible in the wrong one, and the same injection in someone who relies on the frontalis to hold heavy lids can drop the brow instead of lifting it. Part of a crow's-feet session, never a separate purchase.

Platelet-rich plasma injected at the outer eye delivers growth factors to thin skin, and the honest data are mixed. The first objective pilot measured crow's feet by volume, surface, depth and visibility index and found no improvement at three months, though dark-circle colour changed (Mehryan 2014); an uncontrolled study of two monthly sessions in 20 people rated the result good or very good in 45% and mild in 50% (uncontrolled study); a photothermally modulated PRP improved surface roughness and depth measures for at least six months in a whole-face comparison written up with the system's maker (2026 study). Reasonable as a skin-quality add-on to needling; not a treatment for a fold. The regenerative guide covers the systems.

The idea of a toxin that crosses skin without a needle was tested seriously: a topical botulinum toxin gel for lateral canthal lines went through registered trials (registered phase 2 trial) and did not reach the market, and later work compares delivery methods in small clinical studies rather than products (30-woman comparison). The creams sold as "topical Botox" contain peptides, not toxin, and cannot reach a muscle beneath the dermis. The needle remains the route.