Crow's Feet: What the Trials Say About Dose and Duration

Direct Answer

For lines that appear only when you smile, botulinum toxin is the established treatment, and the crow's-feet-specific trials are unusually clear: 12 units per side was the dose a four-arm dose-response RCT settled on, and two multicentre trials put median duration at 119 to 148 days — four to five months, not three. For lines visible at rest, toxin is only part of the answer.

The context worth having first: Cochrane's review of botulinum toxin for facial wrinkles covered 65 randomised trials and 14,919 participants, but 43 of those trials treated the glabella and only seven treated crow's feet. The efficacy figures most often quoted come from the glabellar evidence. Crow's feet have their own smaller, more specific literature — and reading it changes three things: the unit count you should expect, how long it should last, and which side effects actually apply to this area.

1. What crow's feet actually are

The lines fanning out from the outer corner of the eye have a clinical name — lateral canthal lines — and a single mechanical cause. The orbicularis oculi is a ring of muscle around the eye socket, and every time you smile, squint or laugh, its outer fibres pull the skin into folds perpendicular to the direction of pull. Do that a few hundred thousand times against skin that is thin to begin with and progressively loses elastin, and the fold stops unfolding.

Dynamic vs Static — the Distinction That Decides Your Treatment

	Dynamic lines	Static lines
When you see them	Only at maximum smile	At rest, in a neutral expression
What is causing them	Muscle contraction folding the skin	Structural change in the dermis after years of folding, plus photoageing
What responds	Botulinum toxin, well and predictably	Resurfacing, retinoids, sun protection — toxin only partially
Realistic outcome	Near-complete softening for months at a time	Gradual improvement over months, never a reset

Test yourself in a mirror before you read any further. Neutral face, no expression. If the lines are gone, they are dynamic and toxin is the right tool. If they are still there, part of your problem is structural and toxin alone will disappoint you — the trials themselves report separate outcomes for dynamic and static lines, and the static numbers are the softer ones.

2. The evidence gap nobody mentions

Botulinum toxin type A is the most-studied cosmetic intervention in existence, and the Cochrane review of it for facial wrinkles — 65 randomised trials, 14,919 participants — is as good a summary as the field has. Read its breakdown of which region was actually treated, though, and something worth knowing emerges.

Of those 65 trials, 43 treated the glabella and only seven treated crow's feet. The headline efficacy figures people quote — that toxin beats placebo by an enormous margin at week four — are drawn from the review's main comparisons, which are based on single-cycle glabella studies. That does not make toxin ineffective for crow's feet; separate trials establish that it works. It means the number you were quoted probably came from a different muscle group, and the honest position is that the lateral canthal evidence base is roughly one-sixth the size of the glabellar one.

The Cochrane authors' own conclusion is deliberately plain: toxin reduces wrinkles within four weeks of treatment, and probably increases the risk of eyelid ptosis. They also flag what is missing from the whole literature — evidence on multiple treatment cycles, on the frequency of major adverse events, and on duration of effect. Anyone telling you a confident number for how long your third round will last is going beyond what has been published.

3. Dose and duration: what the crow's-feet trials actually found

The good news is that this is one area where the specific studies exist and are unusually clear.

Dose. A double-blind, randomised, placebo-controlled dose-response study compared 3, 6, 12 and 18 units per side against placebo. Every dose beat placebo, higher doses lasted longer and worked harder — but 18 units did not clearly separate from 12, and the authors concluded that 12 units per side is the appropriate dose. That is a useful benchmark to hold when a clinic quotes you a unit count.

Duration. Two multicentre randomised trials followed responders who received 24 units in the crow's feet area. Median duration of response was 125 to 148 days for dynamic lines and static lines in the longer trial, and 119 to 121 days in the second. In other words: four to five months, not three. The three-month figure that circulates online is closer to the glabella experience than to the published crow's-feet data.

What the Trials Support, Numerically

Question	What the published data says	Source strength
How many units per side?	12 units, per a dose-response RCT that tested four doses; 18 units gave no clear additional benefit	Single dedicated dose-ranging trial
How long does it last?	Median 119–148 days across two multicentre RCTs of 24-unit treatment — four months or more	Two RCTs, 833 treated subjects
Does treating the whole upper face last longer?	A meta-analysis of 14 RCTs covering 8,369 subjects found the approved 64-unit cumulative dose across glabella, crow's feet and forehead took a median 182 days to return to non-responder status	Meta-analysis with trial sequential analysis
Does it work on static lines too?	Partially. The duration trials report static-line outcomes separately, and they are real but weaker than the dynamic-line results	Same two RCTs

The upper-face finding is the practical one. If you treat the glabella, crow's feet and forehead together at the approved cumulative dose, the meta-analysis puts median time to non-responder status at about six months rather than four. That reframes the cost conversation: three areas at once is not three times the problem, and spacing single areas across the year may be the more expensive way to buy the same result.

Illustration contrasting lateral canthal lines visible only at full smile with lines that remain visible at rest

4. Safety, and one finding that matters in Asia

The largest controlled safety analysis of onabotulinumtoxinA for facial lines pooled individual patient data from 1,678 participants across nine studies — six randomised, double-blind and placebo-controlled, three open-label. Two of the double-blind studies were specifically for lateral canthal lines at 3 to 18 units per side.

Pooled Safety Findings, 1,678 Participants

Finding	Detail
Overall adverse event rate	Did not differ significantly between toxin and placebo in the double-blind studies
Eyelid sensory disorder	2.5% vs 0.3% for placebo — described by participants as tightness, pressure, heaviness or a feeling of droopiness. Present only in the glabellar studies
Eyelid ptosis	1.8% vs 0% — again, only in the glabellar studies
Treatment-related events overall	24% vs 16% for placebo
Across repeat cycles	Incidence of all three of those events fell significantly as the number of treatment cycles increased
Ethnicity	The cohort was 52% Asian and 43% non-Hispanic white. Eyelid sensory disorder and eyelid oedema were more common in Asian participants

Two things follow from that. The ptosis risk that dominates every consent form is a glabellar phenomenon in this dataset, not a crow's-feet one — which is reassuring if crow's feet are all you are treating, and relevant if you are treating both. And the ethnicity finding is worth raising with a Japanese injector directly: eyelid heaviness and oedema were reported more often in Asian participants, so if you have a lid anatomy that already sits heavy, say so at the consultation rather than after.

The one genuine functional risk of over-treating this area. The orbicularis oculi is not decoration — it closes the eye and drives the tear pump. Excessive or badly placed toxin laterally can produce dry eye, a changed smile, or a smoothed eye area that reads as expressionless in exactly the region where people look for warmth. This is the treatment where under-dosing on the first round is genuinely the better strategy. You can add at the two-week review; you cannot subtract.

5. The static half of the problem

Toxin will not resurface skin. For the lines that are visible at rest, the interventions are the ones that change the dermis, and they work on a scale of months.

For Lines Visible at Rest

Option	What the evidence looks like	Honest expectation
Topical tretinoin	A systematic review of randomised controlled trials of tretinoin for photoageing is the reference point — the best-evidenced topical in dermatology	Slow, cumulative, irritating at the start. The eyelid area needs the lowest strength and the shortest contact time
Non-ablative fractional laser	Multiple randomised periorbital trials exist, including split-face and comparative designs against focused ultrasound	Several sessions, real texture improvement, some downtime. Sun protection is not optional afterwards
Sun protection	The intervention with the largest effect on the rate at which new static lines form, and the least interesting to sell	Prevention, not correction. Start now regardless of what else you do
Filler in the crow's feet themselves	Not a mainstream indication — the skin here is too thin and the area too mobile	Be sceptical of an offer to fill this specific area

The sequencing that makes sense: toxin first, because it removes the folding force and lets everything else work on a surface that is not being creased forty times an hour. Resurfacing second, once you can see what remains at rest.

6. Crow's feet treatment in Japan

Japan has a regulatory quirk worth understanding before you compare prices. The Ministry of Health, Labour and Welfare has approved exactly one botulinum toxin product for cosmetic use — Allergan's Botox Vista. Korean-made toxins are widely used in Japanese clinics and are not MHLW-approved for this indication, which is legal under private-pay practice but is the entire explanation for the price spread you will see.

Crow's Feet (目尻) — Japan Price Tiers

Tier	Price per area	What you are getting
Budget chain	¥4,000–8,000	Almost certainly a Korean-made toxin. Same active molecule, different product, no MHLW cosmetic approval
Mid-range	¥15,000–22,000	Often Xeomin/Bocouture or entry-level Allergan pricing
Premium	¥25,000–35,000	Botox Vista by name, plus a longer consultation
Treatment fee	¥2,000–5,000	Charged separately at many premium clinics — ask whether the quote includes it

If the product matters to you, the phrase to use is "ボトックスビスタ" (Botox Vista) — asking for "Botox" generically will get you whatever the clinic stocks. Our detailed pages cover Botox in Japan, the brands actually in use, what it costs and how that compares in Japan versus the US. For same-day availability, walk-in Botox in Tokyo is the practical guide.

Ask for the unit count, not just the price. "Crow's feet, ¥6,000" tells you nothing about whether you are getting 6 units per side or 12. The dose-ranging trial found a clear dose-response relationship in both magnitude and duration, so a cheap treatment that is also a small treatment is not the bargain it looks like on the price list. Compare yen per unit, and compare that against the 12-units-per-side benchmark.

Chart comparing budget, mid-range and premium price tiers for crow's feet botulinum toxin treatment in Japan

7. Our position

Crow's feet are the treatment where the gap between what is proven and what is promised is unusually narrow — which is refreshing. Toxin works for the dynamic component, the appropriate dose has been established in a dedicated trial, and the duration data says four to five months rather than the three you will hear quoted.

What we would push back on is the framing. Crow's feet are the lines a face makes when it is happy, and the anatomy that produces them is the same anatomy that makes a smile legible. The published complication data is reassuring for this area specifically; the aesthetic risk of over-treatment is not something a trial measures. Start under-dosed, review at two weeks, and treat the static component with resurfacing and sun protection rather than by adding units.

And if you are choosing where to spend a limited budget across the upper face: the meta-analysis showing about six months of effect from treating all three upper-face areas at the approved cumulative dose is the single most useful economic fact on this page.

8. FAQ

How many units of Botox do you need for crow's feet?

A double-blind, randomised, placebo-controlled dose-response study tested 3, 6, 12 and 18 units per side and concluded that 12 units per side is the most appropriate dose. All doses beat placebo and higher doses lasted longer, but 18 units did not clearly separate from 12. Treat that as a benchmark for comparing clinic quotes, not a prescription — your injector decides based on your anatomy.

How long does Botox last in crow's feet?

Longer than the commonly quoted three months. Two multicentre randomised controlled trials of 24 units in the crow's feet area found median response duration of 125 to 148 days in one and 119 to 121 days in the other — four to five months. Separately, a meta-analysis of 14 RCTs found that treating the glabella, crow's feet and forehead together at the approved 64-unit cumulative dose took a median 182 days to return to non-responder status.

Does Botox work on crow's feet you can see when you're not smiling?

Partially. Lines visible at rest are static lines — structural change in the dermis rather than active muscle folding. The duration trials report static-line outcomes separately and they are real but weaker than dynamic-line results. For static lines the interventions that change the skin itself matter: topical retinoids, non-ablative fractional resurfacing, and sun protection.

Is Botox around the eyes risky?

The largest pooled safety analysis — 1,678 participants across nine studies — found overall adverse event rates no different from placebo in the double-blind studies. Notably, eyelid ptosis (1.8%) and eyelid sensory disorder (2.5%) occurred only in the glabellar studies, not the lateral canthal ones. Incidence of those events fell as the number of treatment cycles increased. The Cochrane review still concludes that toxin probably increases the risk of ptosis overall.

Does Botox affect Asian eyelids differently?

The pooled safety analysis is worth knowing about here: its cohort was 52% Asian, and eyelid sensory disorder and eyelid oedema were both more common in Asian participants. That is a fair thing to raise directly at consultation, particularly if your upper lid already sits heavy.

How much does crow's feet Botox cost in Japan?

Budget chains list ¥4,000–8,000 per area, mid-range clinics ¥15,000–22,000 and premium clinics ¥25,000–35,000, with a separate treatment fee of ¥2,000–5,000 at many premium clinics. The spread is mostly product: Japan's MHLW has approved only Allergan's Botox Vista for cosmetic use, so the cheapest tiers are generally Korean-made toxins.

Can filler be used for crow's feet?

It is not a mainstream indication. The skin at the lateral canthus is thin and the area is highly mobile, which is a poor combination for filler. Treat an offer to fill this specific area as a reason to ask more questions.

What's the best order to treat crow's feet?

Toxin first, because it removes the folding force and lets everything else work on skin that is not being creased repeatedly. Then reassess at rest after a few weeks and treat whatever static component remains with resurfacing and retinoids. Sun protection runs throughout — it is the only part of this that prevents new lines rather than treating existing ones.

Sources & Method

Every clinical figure on this page is sourced to a named, PubMed-indexed study, and each PMID below was confirmed against the PubMed record rather than taken from a secondary summary. Where we give a duration or a dose, we state the trial design and the number of participants it came from. The observation about how few Cochrane trials treated crow's feet comes from the review's own description of its included studies. Japanese prices are the ranges verified for our Botox pricing guides in March 2026 and cross-checked against published clinic price lists; treat them as ranges and confirm with the clinic. The treatment sequencing advice and the caution about under-dosing on a first round are our editorial position, not trial findings.

Peer-reviewed literature

Source	Used for
Botulinum toxin type A for facial wrinkles (PMID 34224576) Camargo et al., Cochrane Database of Systematic Reviews, 2021 · Cochrane review, 65 RCTs / 14,919 participants · accessed 14 Aug 2026	Section 2 in full — the 65-trial, 14,919-participant scope, the breakdown showing 43 glabella trials against seven crow's feet trials, the conclusion that toxin reduces wrinkles within four weeks but probably increases ptosis risk, and the authors' statement that evidence on multiple cycles, adverse event frequency and duration of effect is lacking
Dose-response study of botulinum toxin type A in subjects with crow's feet (PMID 15841623) Lowe et al., Dermatologic Surgery, 2005 · double-blind, randomised, placebo-controlled dose-ranging RCT · accessed 14 Aug 2026	The dose figures in section 3 and the FAQ — 3, 6, 12 and 18 units per side tested, all beating placebo, a dose-dependent effect on magnitude and duration, no clear separation between 18 and 12 units, and the authors' conclusion of 12 units per side
Duration of Clinical Efficacy of OnabotulinumtoxinA in Crow's Feet Lines (PMID 27110893) Baumann et al., Dermatologic Surgery, 2016 · two multicentre RCTs, 833 treated subjects · accessed 14 Aug 2026	The duration figures in section 3 — median 125 and 144 days for dynamic lines and 137 and 148 days for static lines in the first study, 119 to 121 days in the second, and the separate reporting of dynamic and static outcomes referenced in sections 1 and 5
Efficacy and Duration of OnabotulinumtoxinA With 64-Unit Dose Optimization (PMID 36099476) Rahman et al., Aesthetic Surgery Journal, 2023 · systematic review and meta-analysis, 14 RCTs / 8,369 subjects · accessed 14 Aug 2026	The upper-face finding in section 3, the tip box and section 7 — median 182 days to return to non-responder status when glabella, crow's feet and forehead lines are treated simultaneously at the approved 64-unit cumulative dose
Safety and tolerability of onabotulinumtoxinA in facial lines: meta-analysis of individual patient data (PMID 19744746) Brin et al., Journal of the American Academy of Dermatology, 2009 · individual patient data meta-analysis, 1,678 participants · accessed 14 Aug 2026	The whole of section 4 — the nine-study pooled design including two lateral canthal line studies at 3 to 18 units per side, the 2.5% vs 0.3% eyelid sensory disorder and 1.8% vs 0% ptosis rates occurring only in glabellar studies, the 24% vs 16% treatment-related event rate, the fall in incidence across treatment cycles, and the finding that eyelid sensory disorder and oedema were more common in the 52% Asian cohort
Topical tretinoin for treating photoaging: a systematic review of randomized controlled trials (PMID 35620028) Sitohang et al., International Journal of Women's Dermatology, 2022 · systematic review of RCTs · accessed 14 Aug 2026	Section 5's characterisation of topical tretinoin as the best-evidenced topical option for photoageing
1550-nm fractional photothermolysis versus intense focused ultrasound for periorbital wrinkles (PMID 38211707) Koza et al., Journal of the American Academy of Dermatology, 2024 · randomised controlled trial · accessed 14 Aug 2026	Section 5's statement that randomised periorbital resurfacing trials exist, including comparative designs against focused ultrasound

Price sources

Source	Used for
Botox Japan ClinicJapan · our own price research, verified March 2026	The crow's feet price tiers in section 6 (¥4,000–8,000 budget, ¥15,000–22,000 mid-range, ¥25,000–35,000 premium), the separate ¥2,000–5,000 treatment fee, and the MHLW approval status of Botox Vista
Botox Brands Japan ClinicJapan · our own price research, verified March 2026	Section 6's account of which products sit at which price tier in Japanese clinics

ClinicJapan editorial

Source	Used for
The glabella-vs-crow's-feet evidence observation ClinicJapan Research Team · Our reading of the Cochrane review, not a claim the review makes	Section 2. The trial counts are Cochrane's own; the inference that widely quoted efficacy figures therefore rest on glabellar data is ours, and it is an argument about where numbers come from, not a claim that toxin fails on crow's feet
Under-dose first, review at two weeks ClinicJapan Research Team · Editorial position, not a trial finding	Sections 4 and 7. No cited study tests this strategy. It follows from the dose-response relationship plus the fact that toxin cannot be reversed, and it is a judgement about aesthetic risk that clinical trials do not measure
Treatment sequencing ClinicJapan Research Team · Editorial method	Sections 5 and 7. Treating dynamic lines before resurfacing static ones is our recommended order; it is a practical inference rather than a comparison any trial has run
Yen-per-unit comparison advice ClinicJapan Research Team · Editorial method	The tip box in section 6. Clinics price by area rather than by unit, so the comparison method is ours

About this guide: This is an independent guide. We receive no commission from Allergan, any toxin manufacturer, any clinic, or any brand named or linked above, and nobody paid to be included. Botulinum toxin is a prescription medical treatment that must be administered by a qualified physician; unit counts and product selection are clinical decisions and the figures here are benchmarks for comparison, not recommendations for your face. Prices change and vary by clinic. Regulatory approval status differs by country and can change. Informational only; not medical advice, and not a substitute for consultation with a doctor.

Crow's Feet: Dose, Duration and What Actually Works