How Long Does Botox Last? What the Trial Data Shows, and When You Can Rebook

By Lucas, Founder & Lead Researcher

Reviewed for medical accuracy against our Medical Review Board  ·  Last updated 26 August 2026

Part of Botox Guides: Basics, Dosing and Safety

The label says the effect on frown lines lasts approximately three to four months, and that is the honest headline. The trial data underneath it is more useful: 80 percent of people met the response threshold at day 30, 70 percent at day 60, 48 percent at day 90 and 25 percent at day 120. So at the three month mark, when most people book their next appointment, fewer than half were still responding.

That is not a criticism of the drug. It is the difference between "when does it stop working" and "when does it stop being measurable", and the gap between those two is where most of the confusion about Botox duration lives.

1. The decay curve, day by day

The two pivotal glabellar line studies enrolled 405 people on BOTOX Cosmetic and 132 on placebo, each receiving a single 20 unit treatment across five sites. A responder was someone the investigator graded as having no or only mild lines at maximum frown. Here is what happened over the following four months:

Notice also how flat the placebo column is: 6%, 3%, 2%, 2%, 2%. Unlike some of the injectable trials on this site, there is almost no placebo effect to subtract here. Whatever you are seeing at day 30 is the drug.

2. You will rate it higher than your injector does

The same trials asked the subjects themselves, using a different question: had their glabellar lines improved at least moderately?

Patients are consistently more positive, and the gap widens as the effect fades: 9 points at day 30, 15 at day 90, 14 at day 120. Be careful what you conclude from that. The investigator was asked whether the lines are objectively gone; you were asked whether things improved. Those are not the same bar, so this is not evidence that patients are fooling themselves. What it does mean is that a clinic quoting patient satisfaction figures and a clinic quoting clinical response rates are quoting different things, and the patient number will always look better.

3. How long Botox takes to work, which is not the same question

People asking how long Botox takes to work usually mean one of three different things: when it starts, when it looks like the result, and when it peaks. Roughly two days, roughly two weeks, and somewhere in the first month.

From the label: chemical denervation of the injected muscle begins one to two days after injection and increases in intensity during the first week. That matches the trial data, where 74 percent were already responders at day 7.

So the useful mental model is not a plateau. It is a ramp of about a week, a peak somewhere in the first month, a long slow decline through months two and three, and very little left by month four. Judging your result before day 14 is judging it on the way up. Our main Botox guide covers the full timeline.

4. Age changes the answer more than anything else

Buried in the same label is a split by age at day 30 that changes the picture considerably:

That is a very large difference, and it needs a caveat stated as plainly as the number itself: only 23 subjects were 65 or over, against 382 under 65. Nine responders out of 23 is a fragile figure, and the confidence interval the label reports for that group runs from minus 17 to plus 51, which means it does not exclude zero. Treat it as a signal worth knowing rather than a settled fact.

The mechanism is not mysterious, though. Toxin relaxes muscle. It does not fill a crease that has become a permanent fold in the skin, and the older the line the more of it is skin rather than movement. The label also notes that responder rates were higher for women than for men.

5. Why each area seems to give a different answer

Search for how long Botox lasts and you will find wildly different numbers by area. Much of that is not biology, it is which question the trial asked:

Crow's feet at 26.1 percent looks like a failure next to frown lines at 80 percent. It is not. The crow's feet trial required a two grade improvement on a composite of investigator and subject assessment, which is a far harder bar than "the lines are now mild". Anyone comparing those two figures without saying so is comparing a high jump with a hurdle.

6. The three month rule, and where it comes from

The label is explicit:

The safety and effectiveness of dosing with BOTOX Cosmetic more frequently than every 3 months have not been clinically evaluated.

That is not a finding of harm. It is an absence of study. But it explains why a reputable clinic will decline to retreat you at eight weeks even when you can see the effect fading, and it is worth knowing that the reason is regulatory caution rather than an arbitrary policy. In the forehead studies, subjects received three cycles over one year, which works out at the same interval.

The same label section explains which direction the risk actually runs. Neutralizing antibodies are rare at cosmetic doses, and in three lateral canthal line trials covering 916 subjects none developed them. But where the label does offer guidance, it is this: the potential for antibody formation may be minimized by injecting with the lowest effective dose given at the longest feasible intervals, and injections at more frequent intervals or at higher doses may lead to greater incidence of antibody formation. That is worth stating plainly because the advice circulating online often runs the opposite way, telling readers that too small a dose is what provokes antibodies. The label says the reverse on both axes.

7. Touch-ups, and the two that mean different things

A touch-up at day 14 and a touch-up at week 8 are not the same request, and conflating them is how people end up dosed more often than any label supports.

The day 14 touch-up corrects the treatment you already had. One brow sitting higher than the other, a line that softened less than the rest, a few units placed where the muscle turned out to be stronger than it looked. This is ordinary work. Section 3 explains why two weeks is the right moment to assess: before then you are judging a result that is still arriving, and asking for corrections at day four produces overcorrection roughly as often as it fixes anything. Many injectors include this in the treatment price. Ask at booking rather than afterwards.

The week 8 top-up is a different thing wearing the same word. Here nothing went wrong: the effect is fading on schedule, and you would like it not to. The BOTOX Cosmetic label states that dosing more frequently than every three months has not been clinically evaluated, and section 6 covers why that interval exists and what it is protecting against. So a clinic that will re-dose you at week eight is not giving you better service. It is going past the point where anyone has evidence, on a drug where the labelled advice is the lowest effective dose at the longest feasible interval.

If you are consistently reaching for a top-up at week eight, the honest read is usually that your dose is too low for your muscle rather than that your interval should be shorter. That is a conversation about units, and section 8 covers what actually changes duration.

8. What actually makes it last longer

• Dose, within reason. An underdosed area fades sooner because there was less to lose. This is the most common reason a result feels short. See the units guide for the labelled figures by area.
• Which area. Large strong muscles such as the masseter hold a result longer than the delicate orbicularis around the eye, which is thin, constantly moving and close to the surface.
• Consistency over time. Regular treatment tends to mean the muscle is never back to full strength when the next dose arrives, so people often report needing less over the years.
• Not aftercare. This is the honest part. The rules about exercise, heat and lying down do not appear anywhere in the label, as covered in the aftercare guide. Following them is cheap and sensible; expecting them to extend your result by weeks is not supported.
• Zinc, and the 30 percent claim. Search for ways to make a result last longer and you will be told to take 50 mg of zinc. That number comes from a single 2012 crossover pilot in 77 patients, and three things about it rarely travel with the figure. The arm that worked was zinc citrate plus phytase, an enzyme taken alongside it; the arm given 10 mg of zinc on its own showed no effect, so taking zinc is not in fact what was tested. The paper also has a published rebuttal, which found no evidence the subjects were zinc deficient to begin with, ambiguity about the real dose and design, unmasking of the crossover, and a potential financial conflict of interest. Its conclusion was that a high level of clinical and scientific skepticism is warranted. And then there is the dose. The Tolerable Upper Intake Level for zinc is 40 mg a day for adults, and the NIH states that doses of 50 mg or more over a period of weeks can inhibit copper absorption, reduce immune function and lower HDL cholesterol. So the protocol circulating on med spa sites, 50 mg daily and ongoing, sits above the ceiling and is taken indefinitely, on the strength of one disputed pilot. If you want to try it, that is a conversation with a doctor who knows what else you take, not a supplement aisle decision.

9. If it stopped working, and what resistance actually means

The phrase people search is Botox resistance, and the worry behind it is that the body has learned to neutralise the drug. That does happen. It is also, at cosmetic doses, close to the least likely explanation for a disappointing result.

Section 6 covers what the BOTOX Cosmetic label reports for its own trials. The wider picture across the class is more useful, because it shows where the risk actually sits. Each of these labels carries the same warning first: the assays differ between products, so these numbers are not a league table.

• Dysport, glabellar lines. 1,554 subjects tested across up to nine treatment cycles. Two were positive for binding antibodies before their first treatment and three more became positive afterwards. None tested positive for neutralizing antibodies.
• Jeuveau. Among 1,414 subjects, two had pre-existing antibodies and two had treatment-emergent antibodies.
• Xeomin. Of 2,649 patients across its trials, 9 (0.3%) were positive for neutralizing antibodies with an unknown baseline status and 4 more (0.2%) developed them after treatment. The label adds that no patient showed a secondary lack of treatment response because of them.

Now the same molecules at medical doses, from the therapeutic labels rather than the cosmetic ones. Dysport reports about 3% in cervical dystonia, and 3.6% developing neutralizing antibodies across 281 long-term spasticity subjects. Botox reports 4 of 326 cervical dystonia patients (1.2%), 2 of 380 adults treated for upper limb spasticity (0.5%), and 2 of 264 children treated for lower limb spasticity (0.8%).

Zero out of 1,554 for glabellar Dysport, against 3.6% for spasticity Dysport. The variable is not the brand. It is the dose. A glabellar cosmetic treatment is 20 units. Cervical dystonia and limb spasticity run into the hundreds. That is also why the label's own advice is the lowest effective dose at the longest feasible interval, which is the same guidance behind the three month rule in section 6.

One more thing the labels say that clinic pages tend to leave out: antibodies do not reliably stop the drug working. Dysport states that in the presence of binding and neutralizing antibodies some patients continue to experience clinical benefit. Botox states that both children who developed neutralizing antibodies continued to experience clinical benefit after subsequent treatment.

So if yours stopped working, the ranked list of likely reasons looks nothing like the one you will find on a forum. Start with the dose you were given in units, then the day you judged it on, then what you were comparing it to. A first treatment is measured against a face that has never been treated, and no later result gets that comparison again. Section 7 covers dose, section 3 covers which day to judge on, and section 2 covers the comparison problem. Antibodies come a long way down.

On tolerance and immunity, which are the other two words people reach for: neither describes a documented mechanism here. What the labels measure is neutralizing antibodies, and they are measured rather than inferred. If you genuinely suspect it after several cycles of no response at an adequate dose, that is a conversation with your injector about dose and record keeping, and one where switching product is a reasonable thing to raise.

10. What repeated use does over years

The question underneath this one is usually whether years of treatment change your face permanently, in either direction. The labels answer part of it and are silent on the rest, and the silence is worth knowing about.

What the mechanism implies. Section 12.1 of the label describes three things that follow repeated treatment: the muscle may atrophy, axonal sprouting may occur with new receptors developing, and reinnervation may occur, slowly reversing the denervation. So the process is described as reversible, and the muscle shrinking from disuse is expected rather than a complication. That is the same mechanism that makes long-term masseter treatment visibly slim a jaw.

Two practical effects people report follow directly from atrophy. Some find they need fewer units over time, because a smaller muscle needs less. Some find that lines look better even before their next treatment, because a muscle used less for years folds the skin less. Neither is a labelled claim.

What the evidence actually covers. This is the part that is usually overstated in both directions. The longest published repeat-dosing data for cosmetic use is not measured in decades:

• Dysport tested 1,554 glabellar subjects across up to nine treatment cycles, which at the labelled interval is a little over two years, and found no neutralizing antibodies at all.
• Botox Cosmetic analysed 916 subjects across three lateral canthal line trials: 1.5% developed binding antibodies and none developed neutralizing antibodies.

That is reassuring within its range and it does not extend past it. Nobody has published a thirty-year cosmetic cohort, and a clinic telling you decades of use are proven safe is going beyond what exists. What can be said is that across the largest repeat-dosing datasets available, the thing people fear most, losing response through immunity, did not happen at cosmetic doses. Section 9 covers that in full.

The one piece of long-term guidance the label does give is about how to treat rather than whether: the lowest effective dose at the longest feasible interval. Over ten years that is the whole of the advice, and section 6 covers why.

11. How the other toxins compare

Each product's label states its own duration and they are not interchangeable claims. Ours are covered separately: how long Dysport lasts, where the label offers up to four months and says nothing at all about onset, and how long Xeomin lasts, where the label gives a median onset of two to seven days and a duration of up to 12 to 16 weeks, which is the most specific of the three.

Daxxify is the outlier, marketed on a longer duration than the others. Whether the extra weeks are worth the price difference is a cost question rather than a duration one, and our price guide has the arithmetic.

Frequently asked questions

Written by

Lucas Pradella

Founder & Lead Researcher, BestBotoxClinics.com

Lucas founded BestBotoxClinics.com to bring clarity to the aesthetics industry, bringing 13+ years of experience in digital research and analysing online data for signals of quality and trust. He is not a physician, his work is research, data analysis and verifying provider credentials.

More about Lucas  ·  How we work

Medical accuracy: BestBotoxClinics.com maintains a Medical Review Board of licensed clinicians. This article is educational and is not medical advice.

Know the count before the quote. The unit count is half the bill, and a low per-unit rate recovered by a high unit count is the commonest way a deal is not one. How to tell.

Deciding which area, not how many units? Botox by facial area covers what each zone actually achieves, which four are FDA approved, and the things Botox cannot fix at all.

Keep reading

More from this guide

Botox Education14 Botox Myths and Misconceptions Botox Comparison GuidesBaby Botox vs. Regular Botox Botox Treatment AreasBotox by Facial Area Botox Comparison GuidesBotox Cosmetic vs. Medical Botox Botox CostsBotox Deals and Discounts Other NeurotoxinsHow Long Does Jeuveau Last Other NeurotoxinsWhich Neurotoxin Lasts Longest? Areas and SafetyIs Botox Bad for You?