Ranked by evidence
Best supplements for migraine prevention
No supplement for migraine earns better than a moderate grade, and the ones that reach it cut attacks modestly rather than stopping them. Everything below is ranked by how strong the evidence is for migraine prevention, not by how popular the product is. All of them need daily use for about three months before you can judge them.
Picks are ranked by the strength of published evidence for this goal in otherwise healthy adults. A persistent symptom needs a clinician's evaluation, and no supplement on this list replaces treatment.
Evidence last reviewed: September 2026
Solid options
| Option | Evidence | Best for | The honest note | |
|---|---|---|---|---|
| Vitamin B2 (Riboflavin) | Moderate evidence | fewer migraine days in adults | At 400 mg a day for 3 months, riboflavin cuts migraine days by roughly 1–2 per month in meta-analyses. Very safe, turns urine bright yellow, and the effect does not carry over to children. | View |
| Migraine Support Formulas | Moderate evidence | riboflavin, magnesium and CoQ10 together | Riboflavin, magnesium and coenzyme Q10 each have meta-analysis support for prevention. Only worth it if the label matches the studied doses: 400 mg riboflavin, 400–600 mg elemental magnesium and 300 mg CoQ10. | View |
| Riboflavin-5-Phosphate | Moderate evidence | only at the full studied dose | The phosphorylated form of B2. The migraine trials used plain riboflavin at 400 mg a day, no trial shows this form works better, and most products supply only 25–100 mg. | View |
Situational
| Option | Evidence | Best for | The honest note | |
|---|---|---|---|---|
| Butterbur | Moderate evidence | under medical supervision only | At 75 mg twice daily it cut attacks by 48% versus 26% on placebo in a 245-patient trial. More than 40 liver injury cases, even with PA-free extracts, mean certified products and liver monitoring or nothing. | View |
Popular for this goal — but the evidence says skip
- Feverfew Limited evidence
Its largest rigorous trial found only 0.6 fewer attacks a month than placebo, and Cochrane rates the evidence low quality.
- Magnesium Oxide Moderate evidence
The oxide-specific migraine trials are small single studies without consistent meta-analytic support, and at about 4% absorption it is an odd choice if magnesium is the point.
- 5-HTP Limited evidence
The migraine trials are small, old and often unblinded, and combining it with triptans risks serotonin syndrome.
- Diamine Oxidase (DAO) Limited evidence
One small trial in patients preselected for low DAO found shorter headaches, and it has not been replicated.
- Ligusticum (Chuan Xiong) Limited evidence
Every migraine trial tested a multi-herb Chinese formula, so the capsules sold as supplements have no trial of their own.
- Gastrodia (Tian Ma) Limited evidence
The gastrodin trials are small, at high risk of bias, mostly compared against active drugs rather than placebo, and often given by injection.
Our verdict
High-dose riboflavin at 400 mg a day is the cheapest and safest place to start; give it three full months. Butterbur has the most striking single trial but a record of liver injury, so it belongs only under medical supervision. Skip feverfew, 5-HTP and magnesium oxide.