General
Botox for Crow's Feet Reliably Softens Smile Lines. Whether Add-On Fillers Help Rests on One Small Trial.

In 1624 the Dutch painter Frans Hals portrayed a young man in an extravagant embroidered doublet, one arm planted on his hip, a waxed moustache turned up at both ends. The picture, now in the Wallace Collection in London, is known as The Laughing Cavalier. The title was attached long after the painter's death, and the sitter is not laughing. He wears the faintest smile, and the name has done the rest of the work. The painting is in the public domain.
Patients arrive with the opposite complaint. They laugh, and the evidence stays on their faces. They ask about Botox for crow's feet, also called treatment for lateral canthal lines or smile lines at the outer corner of the eye. Botulinum toxin is a purified protein that temporarily blocks the signal telling a muscle to contract. For crow's feet it is injected into the muscle that squeezes the eye shut when you smile, the orbicularis oculi. In the trials so far, botulinum toxin softens crow's feet within a month, and the safety record around the eyes is reassuring. Three papers carry the case: a 2026 systematic review and meta-analysis in Aesthetic Surgery Journal Open Forum, a 2024 safety meta-analysis in JAAD International, and a 2025 randomized trial in Aesthetic Plastic Surgery.
Does botulinum toxin soften crow's feet?
Yes. Pooling three placebo-controlled randomized trials with 704 participants, people given incobotulinumtoxinA were about 11.7 times as likely as people given placebo to improve by at least one grade on the Merz Aesthetics Scale 30 days after treatment (risk ratio 11.68, 95% CI 7.49 to 18.20) (Azam et al., 2026). The participants were mostly women with a mean age between 44.8 and 47.8 years (Azam et al., 2026).
The bigger improvements were even more lopsided. For a gain of at least two grades, the pooled risk ratio was 76.26 (95% CI 10.76 to 540.75), a number with a very wide interval because so few people on placebo improved that much (Azam et al., 2026). Investigators rated 28.53 times as many treated participants as having no or mild lines (95% CI 12.01 to 67.82), and the participants' own ratings agreed, at 20.24 times (95% CI 9.18 to 44.62) (Azam et al., 2026).
The reviewers rated most risk-of-bias domains as low across the three trials, with some concerns about missing outcome data in one of them (Azam et al., 2026). The comparison was against placebo, not against a different toxin, so it answers whether toxin works, not which brand works best.
Is Botox around the eyes safe?
The pooled data show no excess of overall adverse events, and the eyelid complications that patients fear were uncommon. In the 2026 meta-analysis, any adverse event was no more frequent with toxin than with placebo (risk ratio 1.13, 95% CI 0.94 to 1.37), and serious adverse events were reported in 1 of 363 treated participants and 5 of 185 given placebo, none judged treatment-related by investigators (Azam et al., 2026).
A larger dataset points the same way. The JAAD International meta-analysis pooled 18 registration studies of onabotulinumtoxinA in 5,298 participants with lines of the glabella, the lateral canthus or the forehead (Brin et al., 2024). In the crow's feet group of the placebo-controlled studies, 36.5% of 1,324 treated participants and 36.3% of 808 given placebo had any adverse event, and eyelid ptosis was reported in none of the treated participants and in 1 on placebo (Brin et al., 2024).
Across all treated areas, the picture was less neutral. Eyelid ptosis occurred in 1.0% of treated participants against 0.1% on placebo (P = .0024), and brow ptosis in 0.6% against none (P = .0098) (Brin et al., 2024). Those figures combine all treated areas, so they are not specific to crow's feet. They are a reminder that eyelid and brow ptosis are real risks of upper-face toxin treatment.
Does adding collagen or a skin booster to the toxin help?
One small randomized trial says yes. It is too small to settle the question. In a Chinese single-centre trial of 40 adults with moderate or severe periorbital wrinkles, 20 received onabotulinumtoxinA alone (24 units in total, 3 sites per side) and 20 received the same toxin plus injections of non-cross-linked bovine collagen into the dermis (Liu & Wang, 2025). Both groups had a total effective rate of 100% at follow-up points out to 6 months (Liu & Wang, 2025).
The difference lay in how good the result was. The abstract reports a markedly effective rate of 90% in the combination group against 40% with toxin alone (P < .01), and great satisfaction in 95% against 20% (P < .01) (Liu & Wang, 2025). No adverse events were reported (Liu & Wang, 2025). The paper's results text gives slightly different control-group figures (42% and 21%) from its abstract, which is one more reason to hold these numbers loosely.
The strongest objection
Some will say the toxin evidence is so consistent that the details hardly matter. That is partly true for the core question. Toxin works on crow's feet, and the effect is large.
But every efficacy number above is a short-term number. The primary endpoint in the meta-analysis was day 30 (Azam et al., 2026). The meta-analysis had only three trials and one product, incobotulinumtoxinA, and the authors name small samples and short follow-up among the limits of the included studies (Azam et al., 2026). The JAAD International analysis is retrospective and ad hoc, and was funded by the toxin manufacturer, whose employees are among the authors (Brin et al., 2024). The add-on trial is a single centre of 40 people, and it did not describe blinding (Liu & Wang, 2025). The evidence supports the injection. It does not yet say how long the smoothing lasts, or whether the extras are worth it.
What should you ask before crow's feet treatment?
Ask which muscle is being treated and why the injection sites were chosen. Crow's feet come from the muscle around the eye, and the trials above treated the lateral part of it (Liu & Wang, 2025). The side effects that matter most around the eye, such as eyelid and brow ptosis, depend on where the toxin spreads (Brin et al., 2024).
Ask whether you will be assessed at rest and when smiling, since the lines are defined by movement. Then ask whether an add-on is being proposed and what it is based on. A recommendation to combine toxin with a collagen or skin-quality injection rests on one 40-person trial (Liu & Wang, 2025).
The Clinical Bottom Line
Botulinum toxin is a well-supported treatment for crow's feet. A pooled analysis of placebo-controlled trials showed large improvements at 30 days (Azam et al., 2026), and the safety data from thousands of participants showed no excess of overall adverse events in the crow's feet group (Brin et al., 2024).
The caution is what the studies leave out. Follow-up was short, the add-on evidence is a single small trial (Liu & Wang, 2025), and the larger safety dataset was funded by the manufacturer (Brin et al., 2024). Expect a temporary result that needs a plan for repeat visits.
The cavalier in the Hals portrait never laughs, and the name stuck anyway. Crow's feet are the opposite: the face laughs, the lines record it, and the injection gives that record a quieter edge.
For clinical assessments and consultation in Seoul: itsdrbock.com
References
Azam, M. T., Azam, M. H., Azam, M. H., Ashraf, S., Khan, M. A., & Aminpoor, H. (2026). Efficacy and safety of incobotulinumtoxinA for the simultaneous treatment of upper facial lines: A GRADE assessed systematic review and meta-analysis. Aesthetic Surgery Journal Open Forum, 8, Article ojag085. https://doi.org/10.1093/asjof/ojag085
Brin, M. F., De Boulle, K., Liew, S., Carruthers, A., Carruthers, J., Rivkin, A., Wu, Y., Kawashima, M., Yushmanova, I., Boodhoo, T. I., & Lee, E. (2024). Safety and tolerability of onabotulinumtoxinA in the treatment of upper facial lines from global registration studies in 5298 participants: A meta-analysis. JAAD International, 14, 4-18. https://doi.org/10.1016/j.jdin.2023.07.021
Liu, Y., & Wang, L. (2025). A randomized controlled trial of botulinum toxin combined with non-cross-linked bovine collagen for improving periorbital wrinkles. Aesthetic Plastic Surgery, 49(24), 6932-6941. https://doi.org/10.1007/s00266-025-04906-9
More From the Blog

Chemical Peels for Acne Work About Equally Well. The Real Choice Is How Your Skin Tolerates the Acid.

Laser Tattoo Removal Works Best on Black and Blue Ink. Whether a Picosecond Laser Does It Better Is Still Debated.
