There are 2 ways to report the best feverfew trial, and they give very different impressions of the same result. Only 1 of them is useful.
Feverfew cut attacks from 4.8 to 2.9 a month. Placebo cut them from 4.8 to 3.5, leaving a difference of 0.6 attacks.
Quick Answer: Does Feverfew Work for Migraines?
Modestly, at best. The largest and most rigorous trial reduced migraine frequency by 1.9 attacks a month while placebo reduced it by 1.3, a difference of 0.6 attacks. A 2025 meta-analysis of 9 trials in 899 participants found frequency down 1.11 and pain severity down 0.63, which was not statistically significant.
Key Takeaways
- Feverfew cut attacks from 4.8 to 2.9 a month in the best trial.
- Placebo cut them from 4.8 to 3.5, a reduction of 1.3 attacks.
- The difference between the 2 arms was 0.6 attacks a month.
- A 2025 pooling of 9 trials found frequency down 1.11 overall.
- Pain severity fell just 0.63, and that was not statistically significant.
Related Products
The Best Single Trial
Cochrane singled this one out as rigorous and larger than the rest, and its numbers are worth laying out in full.
| Measure | Feverfew | Placebo |
|---|---|---|
| Attacks per month at baseline | 4.8 | 4.8 |
| Attacks per month after | 2.9 | 3.5 |
| Reduction | 1.9 | 1.3 |
| Difference between arms | 0.6 attacks per month | — |
| Participants | 218 in that trial | — |
Quoting the 1.9 without the 1.3 is the commonest error in writing about this herb.[1]Feverfew for preventing migraine — Cochrane Database of Systematic Reviews (2015) View source What feverfew is comes first, in our feverfew overview.
What the Preparation Was
The trial that produced that result used a specific standardised extract at a very small amount.
- Preparation: A feverfew CO2 extract, standardised.
- Dose: 6.25 mg, 3 times daily.
- Daily total: 18.75 mg.
- Duration: 16 weeks, after a 4-week baseline.
Responder-rate analysis gave an odds ratio of 3.4 in favour of the extract, and adverse events were lower than placebo at 8.4% against 10.2%.[2]Efficacy and safety of 6.25 mg t.i.d. feverfew CO2-extract (MIG-99) in migraine prevention - a randomized, double-blind, multicentre, placebo-controlled study — Cephalalgia (2005) View source Whether a 1000 mg herb capsule delivers the same thing is answered in which preparation was tested.
The 2025 Pooling
Cochrane could not combine its 6 trials. A 2025 meta-analysis managed 9, and reported both the effects and the disagreement between studies.
| Outcome | Effect | Heterogeneity |
|---|---|---|
| Attack frequency | Down 1.11, p under 0.00001 | 28% |
| Attack duration | Down 4.43, p equals 0.007 | 98% |
| Pain severity | Down 0.63, not significant | 85% |
| Nausea and vomiting | No significant effect | — |
| Photophobia and phonophobia | No significant effect | — |
Heterogeneity of 98% on duration means the included trials disagreed almost completely, so that figure carries far less weight than the frequency result.[3]Systematic review and meta-analysis of Tanacetum parthenium: evaluating its efficacy in migraine relief — Natural Product Research (2025) View source Four associated symptoms moved by nothing at all.
What This Means in Practice
Turning 0.6 attacks a month into a decision is the useful step.
- If you have 8 attacks a month: Expect under 1 fewer.
- Timeframe: 16 weeks before judging.
- What will not change: Nausea, vomiting, light and sound sensitivity.
- What it is not: A treatment for an attack in progress.
For anyone with frequent or worsening migraines, prescription prevention has far larger effect sizes, and that conversation belongs with a clinician. The comparison against other supplements is in how it compares with butterbur.
Running a Fair Trial of It
If the numbers still appeal, the trials give a usable shape.
| Element | What to do |
|---|---|
| Record | Attacks per month, before starting |
| Duration | At least 16 weeks |
| Measure | Frequency, since that is what moved |
| Do not measure | Nausea or light sensitivity; those did not move |
| Stop if | Any red-flag headache feature appears |
A 60-capsule bottle of 1000 mg feverfew covers 2 months of that test at 1 capsule daily.
Seek Urgent Assessment Rather Than Continuing If
- A sudden severe headache unlike any before, or a headache with fever, neck stiffness, confusion, weakness, vision loss or after a head injury
- Any change in the pattern, frequency or character of long-standing migraines
- Pregnancy, or planning a pregnancy
- Taking warfarin, another anticoagulant, or an antiplatelet medicine
- A known allergy to ragweed, chrysanthemums, marigolds or daisies
Frequently Asked Questions
Does feverfew work for migraines? +
Modestly. The best trial produced a difference of 0.6 attacks a month against placebo, and a 2025 pooling of 9 trials found frequency down 1.11.
How much does it reduce attacks? +
In the largest trial, feverfew cut attacks by 1.9 a month and placebo by 1.3. The difference is 0.6 attacks.
How long does it take feverfew to work on migraines? +
The trials ran 16 weeks after a 4-week baseline. Judging it at 2 or 3 weeks is not a fair test.
Does it help migraine nausea? +
No. The 2025 meta-analysis found no significant effect on nausea, vomiting, photophobia or phonophobia, which is 4 outcomes unchanged.
Does it shorten attacks? +
The pooled duration figure was down 4.43, with heterogeneity of 98%. That level of disagreement between trials makes the number unreliable.
Does it reduce pain severity? +
Not significantly. The pooled figure was down 0.63 with a p-value of 0.14.
What is the best way to take feverfew for migraines? +
The positive trial used a standardised CO2 extract at 6.25 mg 3 times daily. Form matters more than the milligram number on the front of a bottle.
Can I take it with my migraine medication? +
Ask first, particularly with anticoagulants. Prescription prevention has far larger effect sizes than 0.6 attacks a month.
Is it worth trying? +
It is cheap and well tolerated, with 8.4% adverse events against 10.2% on placebo. The honest expectation is under 1 fewer attack a month.
What deficiency is linked to migraines? +
Magnesium is the 1 most often raised, and it has its own separate literature. Feverfew is not a nutrient, so there is no feverfew deficiency.
Are any herbs better than feverfew for migraine? +
A 2020 review of 19 studies called feverfew mixed and butterbur positive but limited. Neither of the 2 has a large effect size.
Should I stop if it does nothing? +
After 16 weeks with an unchanged attack count, yes. That is the timeframe the trials used.
Related Reading
Related Products
