Herbs and Supplements for Migraines and Headaches, Ranked by Evidence
The supplements with the best evidence for preventing migraines are riboflavin (vitamin B2) at 400 mg a day, magnesium at 400–600 mg a day, and coenzyme Q10 at 300 mg a day. All three are rated 'probably' or 'possibly' effective in the American Academy of Neurology and American Headache Society (AAN/AHS) 2012 guideline on migraine prevention. Butterbur has the largest single-trial effect of any herb but carries a liver warning serious enough that the guideline withdrew its endorsement. Feverfew, ginger, melatonin and omega-3 fats have smaller but genuine trial support.
This guide is for people with episodic migraine or tension-type headache who want fewer headache days each month. Prevention supplements are not painkillers: most take 8–12 weeks to work, and the realistic win is one to three fewer headache days a month, not zero. For stopping an attack already under way, only ginger (migraine) and topical peppermint oil (tension headache) have acute-use data.
See a doctor rather than self-treating if your headaches are new after age 50, are the worst you have ever had or came on in seconds, come with fever, stiff neck, weakness, confusion or vision loss, wake you from sleep, or if you use painkillers more than 10 days a month. That last pattern is medication-overuse headache, and no supplement fixes it.
Quick answer
- Riboflavin (vitamin B2) — 400 mg/day cut migraine frequency by about half in a 3-month RCT; guideline-endorsed
- Magnesium — 400–600 mg/day reduced attack frequency and intensity in a meta-analysis of 11 trials
- Coenzyme Q10 — 300 mg/day reduced attack frequency and duration in a 6-study meta-analysis
- Butterbur (PA-free only) — 150 mg/day halved attacks in a 245-person RCT, but needs a liver check
- Feverfew — about 0.6 fewer attacks a month in the best trial; Cochrane calls the evidence mixed
- Ginger — 250 mg of powder matched sumatriptan for stopping an attack in a 100-person trial
The 10 best herbs and supplements for migraine and headaches
Ranked by the strength of human evidence. Strong = multiple trials or a meta-analysis. Moderate = a few small trials. Preliminary = early or animal data.
In the 1998 Schoenen trial, 55 migraine patients took 400 mg riboflavin or placebo for 3 months; 59% of the riboflavin group had at least a 50% drop in attacks versus 15% on placebo. Later adult trials were mostly positive and the AAN/AHS 2012 guideline rates it Level B, 'probably effective'. Trials in children using lower doses were negative, so the dose matters.
- Key evidence
- Schoenen 1998 RCT (Neurology); AAN/AHS 2012 guideline, Level B
- Studied dose
- 400 mg/day riboflavin in one or two doses, for a full 3 months
- Best for
- Anyone who wants the safest first-line preventive
- Cautions
- Turns urine bright yellow, which is harmless. Almost no interactions; separate from antibiotics by 2 hours as it may slightly reduce their absorption.
Magnesium
Strong evidenceA 2016 meta-analysis of 11 randomised trials found oral magnesium reduced migraine frequency and intensity, and intravenous magnesium eased acute attacks within 15–45 minutes. The best oral trial (Peikert 1996) used 600 mg of magnesium dicitrate daily for 12 weeks and cut attack frequency by 41.6% versus 15.8% on placebo. The AAN/AHS guideline rates it Level B; it is especially logical in migraine with aura and menstrual migraine, where magnesium tends to run low.
- Key evidence
- Chiu 2016 meta-analysis of 11 RCTs (Pain Physician); AAN/AHS 2012 guideline, Level B
- Studied dose
- 400–600 mg/day elemental magnesium as citrate, glycinate or oxide, 8–12 weeks
- Best for
- Migraine with aura, menstrual migraine, people who also have cramps or constipation
- Cautions
- Loose stools are the main side effect; glycinate and citrate are gentler than oxide. Avoid high doses in kidney disease. Separate from thyroid medication, bisphosphonates and quinolone or tetracycline antibiotics by 2 hours.
Coenzyme Q10
Moderate evidenceSándor's 2005 RCT gave 42 patients 100 mg CoQ10 three times daily or placebo for 3 months; 47.6% of the CoQ10 group had at least 50% fewer attacks versus 14.4% on placebo. A 2021 meta-analysis of 6 trials (371 patients) confirmed shorter and less frequent attacks, though not milder ones. The AAN/AHS guideline rates it Level C, 'possibly effective', and no trial saw results before 4 weeks.
- Key evidence
- Sazali 2021 meta-analysis of 6 RCTs (BMJ Open)
- Studied dose
- 100 mg three times daily (300 mg/day) with food, at least 3 months
- Best for
- People on statins, which lower CoQ10, or who find magnesium upsets their gut
- Cautions
- Can blunt warfarin, so check INR if combined. May modestly lower blood pressure and blood sugar. Mild nausea or insomnia if taken late.
In Lipton's 2004 trial of 245 patients, 75 mg of a PA-free butterbur extract twice daily for 4 months cut attack frequency by 48% versus 26% on placebo; a smaller German trial agreed. The AAN/AHS guideline originally gave it Level A, its highest rating, then withdrew that in 2015 after liver-injury reports and concern about pyrrolizidine alkaloid (PA) contamination in unregulated products. The efficacy is real; so is the product-quality problem.
- Key evidence
- Lipton 2004 RCT of 245 patients (Neurology); AAN endorsement withdrawn 2015
- Studied dose
- 50–75 mg twice daily of a certified PA-free extract standardised to 15% petasins, 4 months
- Best for
- People who failed riboflavin or magnesium and can get a certified PA-free product plus a liver test
- Cautions
- Use only products certified PA-free: raw butterbur and uncertified extracts contain pyrrolizidine alkaloids that damage the liver and are carcinogenic. Avoid in pregnancy, breastfeeding and liver disease; stop for dark urine, pale stools, itching or jaundice. Ragweed-family plant, so it can trigger allergy.
The best trial, Diener 2005, gave 170 patients 6.25 mg of a CO2 extract (MIG-99) three times daily for 16 weeks; attacks fell by 1.9 a month versus 1.3 on placebo, a modest 0.6-attack difference. The 2015 Cochrane review of 6 trials and 561 patients called the evidence mixed and the benefit small; the AAN/AHS guideline rates it Level B. Dried-leaf products vary wildly in parthenolide content, which probably explains the inconsistency.
- Key evidence
- Cochrane 2015 review of 6 RCTs (Wider et al.)
- Studied dose
- 50–150 mg/day dried leaf standardised to 0.2–0.35% parthenolide, or 6.25 mg MIG-99 three times daily, 3–4 months
- Best for
- Frequent mild migraines when you want a plant-based option after the vitamins
- Cautions
- Stopping suddenly after long use can cause rebound headaches and joint aches. Mild antiplatelet effect: caution with warfarin, aspirin and clopidogrel, and stop 2 weeks before surgery. Avoid in pregnancy and daisy-family allergy.
Ginger is the one herb with acute-attack data. In a 2014 Iranian RCT, 100 migraine patients took 250 mg of ginger powder or 50 mg sumatriptan at onset; two hours later, pain relief was statistically similar (about 64% versus 70% responders) with fewer side effects on ginger. A 2019 emergency-department trial found 400 mg of ginger extract added to IV ketoprofen gave greater pain reduction than ketoprofen alone. It also treats the nausea of an attack.
- Key evidence
- Maghbooli 2014 RCT, ginger vs sumatriptan (Phytotherapy Research)
- Studied dose
- 250–500 mg ginger powder or 400 mg extract at the first sign of an attack; repeat once after 2 hours if needed
- Best for
- Acute attacks with nausea; people who cannot take triptans
- Cautions
- Mild blood-thinning effect; caution with warfarin and before surgery. Can cause heartburn. Keep to 1 g/day in pregnancy.
Melatonin
Moderate evidenceA 2016 Brazilian RCT of 178 migraine patients compared 3 mg melatonin, 25 mg amitriptyline and placebo for 3 months; melatonin cut headache days by 2.7 a month versus 2.2 for amitriptyline and 1.1 for placebo, with weight loss instead of amitriptyline's weight gain. An earlier Norwegian trial of 2 mg extended-release was negative, so formulation may matter. Small open studies also report benefit in cluster headache at 10 mg.
- Key evidence
- Gonçalves 2016 RCT, melatonin vs amitriptyline vs placebo (J Neurol Neurosurg Psychiatry)
- Studied dose
- 3 mg immediate-release at bedtime for migraine (10 mg has been used in cluster headache), 3 months
- Best for
- Migraine with poor sleep; people offered amitriptyline who want fewer side effects
- Cautions
- Morning drowsiness; adds to sedatives and alcohol. May interact with blood thinners and immunosuppressants and can affect blood sugar in diabetes. Supervised use only in children.
Omega-3 fats (EPA and DHA)
Moderate evidenceA 2021 BMJ trial randomised 182 people with frequent migraines to a diet high in omega-3 (about 1.5 g EPA+DHA a day), high omega-3 plus low omega-6, or control for 16 weeks. The omega-3 diets reduced headache hours per day and headache days by roughly 1.3–2 a month relative to control. A 2024 RCT of 2 g EPA daily also reduced attack frequency. The effect is modest but fits omega-3's known effect on pain-signalling lipids.
- Key evidence
- Ramsden 2021 RCT of 182 patients (BMJ)
- Studied dose
- 1.5–2 g/day combined EPA+DHA from fish oil, or 2 g EPA, for at least 16 weeks
- Best for
- People who eat little oily fish, or who also have cardiovascular or mood concerns
- Cautions
- Above 3 g/day raises bleeding risk with warfarin, aspirin or clopidogrel; stop 1–2 weeks before surgery. Very high doses have been linked to atrial fibrillation. Fishy burps; refrigerate capsules.
Göbel's 1996 crossover study of 41 patients and 164 tension-type headache episodes found a 10% peppermint oil solution rubbed on the forehead and temples relieved pain as well as 1,000 mg of paracetamol, with onset within 15 minutes. Later work by the same group and a 2019 trial of intranasal peppermint oil for migraine (comparable to lidocaine) support a real, if local, effect through cold receptors and muscle relaxation. This is acute relief, not prevention.
- Key evidence
- Göbel 1996 RCT vs paracetamol (Nervenarzt)
- Studied dose
- 10% peppermint oil in ethanol, or a few drops in a carrier oil, spread across forehead and temples at onset; repeat after 15–30 minutes
- Best for
- Tension-type headache; anyone wanting a drug-free acute option
- Cautions
- Keep away from eyes and broken skin, and never apply near the face of a baby or small child (can cause breathing problems). Do not swallow undiluted oil. Skin irritation in some people.
Low vitamin D is common in migraine clinics, and a 2021 meta-analysis of 5 small RCTs found supplementation reduced attack frequency by about 2 attacks a month compared with placebo, with no effect on severity. The trials used 1,000–4,000 IU a day for 12–24 weeks and mostly enrolled people with low baseline levels. It is the weakest vitamin here and most likely to help if you are deficient.
- Key evidence
- 2021 meta-analysis of 5 RCTs (Hu et al.)
- Studied dose
- 1,000–2,000 IU/day vitamin D3, or a dose chosen after a blood test, for 12 weeks or more
- Best for
- People with a measured low 25(OH)D, darker skin, or little sun
- Cautions
- Above 4,000 IU/day needs a blood test; excess raises calcium and can cause kidney stones. Caution with thiazide diuretics, sarcoidosis and hyperparathyroidism.
How to choose
- Start with riboflavin 400 mg: cheap, nearly side-effect free, guideline-endorsed, and the trial effect is as large as anything else here. Give it 3 months.
- If you have aura, menstrual migraine, cramps or constipation, add or choose magnesium 400–600 mg at night. Many neurologists combine it with riboflavin; the two do not interact.
- If you take a statin, or riboflavin and magnesium did nothing after 3 months, try CoQ10 300 mg/day. Some clinics use all three as a 'migraine stack'; only the parts, not the stack, have been tested.
- If you want the biggest effect size and can get a certified PA-free product and a baseline liver test, butterbur 150 mg/day is the most effective herb in trials. Skip it entirely if you cannot verify PA-free status.
- For stopping an attack: ginger 250–500 mg at onset for migraine, 10% peppermint oil on the temples for tension-type headache. Neither replaces a triptan for severe attacks.
- If sleep is the main problem, melatonin 3 mg at bedtime does double duty. Curcumin is not ranked because its trials are tiny and mostly combined it with omega-3. Ask AskMN if you want a plan checked against your medications.
What to skip
CBD oil
No randomised trial exists for migraine prevention or treatment. The positive reports are surveys and open-label studies, which cannot separate the drug from expectation.
Daith piercing
Based on a loose analogy to acupuncture points, with no controlled study. The one survey found most people's benefit faded within months, the signature of placebo.
Curcumin on its own
Two small Iranian trials reported fewer attacks, but one combined curcumin with omega-3 and the other used a nano-formulation unlike most products. Promising, not proven.
'Migraine detox' teas and cleanses
Migraine is a brain-excitability disorder, not a toxin problem, and dehydration and skipped meals from a cleanse are two of the best-documented triggers.
Safety and when to see a clinician
- See a clinician urgently for a headache that is the worst of your life, starts in seconds, follows a head injury, or comes with fever, stiff neck, confusion, weakness, slurred speech or vision loss. See one soon if headaches are new after 50, steadily worsening, or waking you at night.
- Painkillers (triptans, ibuprofen, paracetamol, codeine) on 10 or more days a month cause medication-overuse headache. No supplement fixes this; the treatment is a supervised break from the painkillers.
- Butterbur must be certified PA-free. Pyrrolizidine alkaloids cause liver damage and cancer, and the risk comes from contaminated products. Avoid it in pregnancy, breastfeeding and liver disease, and get a liver test if you use it for more than a few months.
- Blood thinners: feverfew, ginger, omega-3 above 3 g/day and CoQ10 all interact with warfarin or antiplatelet drugs. Stop feverfew, ginger and fish oil 1–2 weeks before surgery.
- Pregnancy: magnesium and riboflavin are generally considered safe at these doses, which is why obstetric headache clinics favour them. Avoid feverfew and butterbur, keep ginger under 1 g/day, and discuss melatonin and CoQ10 with your midwife or doctor.
- Melatonin adds to sedatives and alcohol; magnesium and vitamin D need dose limits in kidney disease. On five or more medicines, ask a pharmacist to check the whole list.
What matters as much as any herb
- Regular sleep: both too little and too much sleep trigger migraine, and fixing an irregular schedule reduces headache days as much as some preventives.
- Do not skip meals: fasting and low blood sugar are among the most consistent triggers. Eat protein every 4–5 hours on migraine-prone days.
- Steady caffeine and hydration: dehydration triggers attacks, and so does a sudden caffeine drop. Keep caffeine modest and constant, or taper slowly.
- Aerobic exercise: 40 minutes three times a week matched topiramate for prevention in a 2011 Swedish RCT, without the side effects.
- A headache diary: three months of tracking attacks, triggers and painkiller days is the most useful thing you can bring to a doctor, and it tells you honestly whether a supplement is working.
Frequently asked questions
What is the best natural supplement for migraines?
Riboflavin 400 mg a day and magnesium 400–600 mg a day have the best mix of evidence, safety and cost, and both are rated 'probably effective' in the AAN/AHS guideline. CoQ10 300 mg a day is the usual third choice. Butterbur has the largest trial effect but needs a certified PA-free product and liver monitoring.
How long do migraine supplements take to work?
Most prevention trials ran 3 months, with benefit appearing between weeks 4 and 12. If headache days have not changed after 3 months at a full dose, that supplement is unlikely to work for you and it is reasonable to switch.
Which magnesium is best for migraines?
Trials used citrate, oxide and dicitrate, so the salt matters less than reaching 400–600 mg of elemental magnesium. Glycinate and citrate cause fewer loose stools than oxide at those doses; glycinate is the gentlest if your gut is sensitive.
Is butterbur safe for migraines?
Certified PA-free extracts were safe in the two main trials, which ran 4 months. The danger is pyrrolizidine alkaloids in raw butterbur or poorly made extracts, which damage the liver. Use only certified PA-free products, avoid it in pregnancy and liver disease, and ask about a liver test for long-term use.
Can herbs stop a migraine once it has started?
Ginger is the only herb with controlled acute-attack data: 250 mg of ginger powder at onset matched 50 mg sumatriptan at two hours. For tension-type headache, 10% peppermint oil on the forehead and temples relieved pain comparably to 1,000 mg paracetamol. Neither is as reliable as prescription treatment for severe attacks.
Can I take riboflavin, magnesium and CoQ10 together?
Yes; they do not interact, and many headache clinics prescribe the combination. No trial has tested the stack, so expect the sum of modest individual effects rather than something dramatic.
Does feverfew really prevent migraines?
Modestly at best: about 0.6 fewer attacks a month than placebo in the best trial, and the Cochrane review calls the evidence mixed. Standardised extracts did better than dried leaf. It is a third or fourth choice, and stopping suddenly after long use can cause rebound headaches.
References
- Schoenen J, Jacquy J, Lenaerts M. Effectiveness of high-dose riboflavin in migraine prophylaxis. A randomized controlled trial. Neurology (1998)View source
- Holland S, Silberstein SD, Freitag F, et al.. Evidence-based guideline update: NSAIDs and other complementary treatments for episodic migraine prevention in adults. Neurology (2012)View source
- Chiu HY, Yeh TH, Huang YC, Chen PY. Effects of intravenous and oral magnesium on reducing migraine: a meta-analysis of randomized controlled trials. Pain Physician (2016)
- Lipton RB, Göbel H, Einhäupl KM, Wilks K, Mauskop A. Petasites hybridus root (butterbur) is an effective preventive treatment for migraine. Neurology (2004)View source
- Sándor PS, Di Clemente L, Coppola G, et al.. Efficacy of coenzyme Q10 in migraine prophylaxis: a randomized controlled trial. Neurology (2005)
- Sazali S, Badrin S, Norhayati MN, Idris NS. Coenzyme Q10 supplementation for prophylaxis in adult patients with migraine: a meta-analysis. BMJ Open (2021)
- Wider B, Pittler MH, Ernst E. Feverfew for preventing migraine. Cochrane Database of Systematic Reviews (2015)
- Maghbooli M, Golipour F, Moghimi Esfandabadi A, Yousefi M. Comparison between the efficacy of ginger and sumatriptan in the ablative treatment of the common migraine. Phytotherapy Research (2014)
- Gonçalves AL, Martini Ferreira A, Ribeiro RT, et al.. Randomised clinical trial comparing melatonin 3 mg, amitriptyline 25 mg and placebo for migraine prevention. Journal of Neurology, Neurosurgery and Psychiatry (2016)
- Ramsden CE, Zamora D, Faurot KR, et al.. Dietary alteration of n-3 and n-6 fatty acids for headache reduction in adults with migraine: randomized controlled trial. BMJ (2021)
- Göbel H, Fresenius J, Heinze A, Dworschak M, Soyka D. Effectiveness of Oleum menthae piperitae and paracetamol in therapy of headache of the tension type. Nervenarzt (1996)
- NIH Office of Dietary Supplements. Magnesium: Fact Sheet for Health Professionals (2022)View source
Related conditions
More guides
Not sure which one fits you?
Ask AskMN. It knows these studies, checks interactions with the medications you tell it about, and it is free and unlimited, no account needed.
Ask about migraine and headachesThis guide is educational and is not medical advice. Herbs and supplements can interact with prescription medicines and are not regulated like drugs. Talk with a clinician or pharmacist before starting anything new, especially if you are pregnant, breastfeeding, have a chronic condition, or take medication.