Riboflavin (B2)

Best-in-class: Riboflavin 5'-Phosphate

The active form of vitamin B2, a cofactor for cellular energy (FAD/FMN), glutathione recycling, and — notably — migraine prevention.

How Riboflavin (B2) actually works

Riboflavin is the backbone of FAD and FMN, the flavin coenzymes that carry electrons in complex I and complex II of the electron transport chain. It is also required by glutathione reductase to regenerate reduced glutathione, and by MTHFR — which is why riboflavin status modifies the effect of the common MTHFR variant on homocysteine. Riboflavin 5'-phosphate is the pre-activated form.

Where to get Riboflavin (B2)

The evidence for Riboflavin (B2)

Graded by what exists behind each claim.

✅ Clinically validated

• RCTs support high-dose riboflavin (400 mg) for reducing migraine frequency.
• Essential cofactor for energy metabolism and activating B6/folate.

📊 Correlative data

• Low riboflavin status is associated with migraine and impaired methylation.

🧪 Theoretical / extrapolated benefits

• Its role in mitochondrial energy underpins the migraine benefit — mechanistically well-grounded.

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What Riboflavin (B2) actually does

Riboflavin is the only page in this cohort where the evidence comes first and the caveats come second, so it is worth being precise about what the molecule is. Riboflavin is 7,8-dimethyl-10-ribityl-isoalloxazine: a tricyclic isoalloxazine ring that can accept one electron or two, and a ribityl tail that exists so the cell can phosphorylate it. That one-or-two-electron flexibility is the point. NAD and NADH move two electrons at a time; the respiratory chain has to hand electrons to coenzyme Q one at a time; a flavin is the adapter that bridges those two arithmetics, and there is no other cofactor in human biochemistry that does it.

Two enzymes turn the vitamin into the cofactor, and both are named. Riboflavin kinase phosphorylates the ribityl tail to make flavin mononucleotide, FMN. FAD synthase then adenylylates FMN to make flavin adenine dinucleotide, FAD. Nothing downstream uses free riboflavin; every flavoenzyme in the body holds FMN or FAD, and most hold it so tightly that the cofactor is effectively part of the protein.

Where those cofactors sit, in the order that matters for migraine. Complex I of the respiratory chain — NADH:ubiquinone oxidoreductase — carries a single FMN at the tip of its matrix arm, and that FMN is the entry point for every electron the chain receives from NADH. Complex II, succinate dehydrogenase, carries a covalently bound FAD. The electron-transfer flavoprotein and its ubiquinone oxidoreductase, which is how fatty acid beta-oxidation delivers electrons, are both FAD enzymes. Glutathione reductase, the enzyme that regenerates reduced glutathione from its disulfide, is an FAD enzyme — which is why the classic assay of riboflavin status is not a blood level at all but the erythrocyte glutathione reductase activation coefficient, the ratio of the enzyme's activity with added FAD to its activity without. A cofactor-starved enzyme jumps when you give it the cofactor, and the size of the jump is the deficiency.

Two more flavoenzymes that make this vitamin the hinge of the B-vitamin group. Pyridoxine 5'-phosphate oxidase, the enzyme that produces the active form of vitamin B6, is FMN-dependent. Methylenetetrahydrofolate reductase — MTHFR, the enzyme that makes 5-methyltetrahydrofolate and therefore sets the rate at which homocysteine is remethylated — is FAD-dependent. So a riboflavin shortfall degrades B6 activation and folate cycling at the same time, which is the mechanistic reason this molecule appears inside methylation formulas rather than only in energy formulas.

Getting it in is a transporter problem, not a solubility problem. Intestinal uptake runs through the riboflavin transporter family SLC52: RFVT3, the product of SLC52A3, sits on the apical membrane of the enterocyte and is the route by which dietary riboflavin is taken up, with RFVT1 and RFVT2 handling the basolateral and systemic side Yoshimatsu 2014. A carrier is a fixed number of protein molecules, so it saturates, and that single fact governs every dosing decision on this page.

The migraine hypothesis, stated as a hypothesis. If migraine brain runs with a reduced mitochondrial phosphorylation reserve — and phosphorus magnetic resonance spectroscopy work is the origin of that idea — then loading the one cofactor at the entry point of complex I raises flux through the chain and raises the threshold at which a cortical event propagates. That is a mechanism which predicts exactly the trial result it got: a slow onset, because flavoenzyme pools turn over in weeks, and a benefit in attack frequency rather than in the pain of an attack already underway.

Cell, rodent, human — and where it stops

Transporter, in cells and in mouse. RFVT3 was characterized in intestinal preparations and cell models, where knocking it down reduces riboflavin uptake and its expression sits on the apical enterocyte surface Yoshimatsu 2014. This is the step everything else is downstream of.

Pharmacokinetics, in people, and the number that decides the dose. Oral and intravenous riboflavin were compared directly in healthy adults: absorption is carrier-mediated and saturable, so the fraction absorbed falls away steeply as a single dose rises past the high-twenties of milligrams, and what is absorbed above the tissues' capacity to phosphorylate it is filtered and lost in urine within hours Zempleni 1996. Two consequences follow that almost no product page states. A 400 mg dose is not 400 mg of exposure. And splitting the same daily total into two or three doses gets more of it across the gut than swallowing it once, because saturation is per dose, not per day.

The randomized trial, in people, at a pharmacological dose. Fifty-five patients with migraine, 400 mg of riboflavin daily for three months against placebo, with attack frequency as the endpoint: 59% of the riboflavin group improved by at least half against 15% on placebo, and the separation appeared in the third month rather than the first Schoenen 1998. That is a number needed to treat of roughly two and a half, which is better than most migraine prophylactics with a prescription pad behind them. The subsequent systematic review and meta-analysis are more cautious — the trial base is small, adult and pediatric results diverge, and the pediatric trials at 200 mg are largely negative Thompson 2017Chen 2022. A separate randomized, placebo-controlled multicentre trial of riboflavin combined with magnesium and coenzyme Q10 improved migraine symptoms but not the primary attack-frequency endpoint, which is a useful reminder that a combination product is not the same experiment Gaul 2015.

Now the obstacle, and on this page it runs the opposite way to every other vitamin here. The usual objection to a vitamin trial is that it was run in deficient people and you are replete, so your expected effect is zero. That objection does not apply to riboflavin for migraine. The patients in the 400 mg trial were not selected for deficiency and 400 mg is roughly three hundred times the adult reference intake of about 1.3 mg — nobody corrects a deficiency with three hundred times the requirement. This is a drug dose of a vitamin, aimed at a mitochondrial flux problem rather than at a nutrient gap. The real obstacle is the saturable transporter Zempleni 1996Yoshimatsu 2014: the trial's benefit is attached to swallowing 400 mg a day for three months, and nobody has established how much of that dose has to be absorbed for the effect, which means nobody can tell you whether 100 mg split across the day would do the same job for a quarter of the capsules.

Riboflavin (B2) — which form, and does it matter

Riboflavin 5'-phosphate is sold as the active form and it is not absorbed as the active form. Alkaline phosphatase on the intestinal brush border strips the phosphate off in the lumen, because the transporter that carries the vitamin across the enterocyte carries riboflavin, not its phosphate ester Yoshimatsu 2014. Whichever you swallow, free riboflavin is what crosses, and riboflavin kinase inside the cell puts the phosphate back on. The activated form is a marketing distinction that is real chemistry pointed in the wrong direction: the phosphate is added and removed twice on the way to the enzyme.

There is one honest argument for the phosphate ester and it is about solubility, not activation. Riboflavin free base is poorly soluble in water; the sodium salt of the 5'-phosphate is far more soluble, which matters for a liquid, an injection or a capsule that has to dissolve fast. It does not change what reaches complex I.

The dose arithmetic against the label. The filed Supplement Facts panel for the product this site links declares riboflavin as riboflavin 5'-phosphate sodium at 36.5 mg per capsule, one to three capsules daily Office of Dietary Supplements. Follow that suggestion at its maximum and you are taking 110 mg a day. The migraine trial ran at 400 mg Schoenen 1998, which is eleven capsules. This site's own card says up to 400 mg daily for migraine support, and both statements are true at once — the dose is right and the bottle does not conveniently supply it. If migraine is the reason you are buying, the cheap plain-riboflavin 400 mg tablet is the form the trial actually used.

Split it, and take it with food. Absorption saturates per dose Zempleni 1996, and riboflavin uptake is better with a meal because gastric emptying slows and the transporter is presented with a lower concentration for longer. Two hundred milligrams twice daily with meals is, on the pharmacokinetics, a larger absorbed dose than 400 mg once — and that reasoning is labeled extrapolation, because the trial dosed 400 mg once daily and nobody has run the split-dose comparison.

What would have to be true, and how you would know it was not

1. A headache diary with a defined baseline, and the effect arrives late. Record migraine days for 28 days before the first capsule — date, duration in hours, peak severity out of ten, and whether you took an acute medicine — then take 400 mg daily and record for another 84 days. The prediction is a fall of at least half in monthly migraine days in roughly three of every five people who complete it, with placebo-level response in about one in six, and essentially nothing visible in the first month Schoenen 1998. Stopping at four weeks because nothing happened is how most people get a false negative from this molecule.

2. Yellow urine within four hours, which proves absorption and proves nothing else. Fluorescent yellow-green urine is filtered riboflavin above what the tissues could phosphorylate. It is a one-hour confirmation that the capsule dissolved and the transporter worked, and it is not a dose-finding signal — the color appears at 25 mg and at 400 mg alike.

3. Erythrocyte glutathione reductase activation coefficient, if you can get the assay, and homocysteine if you cannot. The activation coefficient falls toward 1.0 as riboflavin status improves, because a saturated flavoenzyme cannot be activated further. Most consumer labs do not run it. The proxy worth ordering instead is fasting total homocysteine, and the prediction is specific: in a carrier of the MTHFR 677 variant with a homocysteine above roughly 12 micromoles per liter, riboflavin alone should lower it by a measurable amount at twelve weeks, because the variant enzyme's defect is FAD binding. In a non-carrier with a normal homocysteine, predict no movement at all. Labeled extrapolation: that comparison has been reported in observational and small interventional work, not in a trial powered for it.

4. The prediction that cuts against the product: nothing measurable will happen to your energy. No change in resting heart rate, no change in a timed cycling effort, no change in how tired you feel at four in the afternoon. Flavoenzyme saturation in a person eating dairy, eggs, meat or fortified cereal is already complete, and a saturated enzyme does not run faster when you add more cofactor. If you do not have migraine, the honest expected value of this bottle is bright urine.

What will fool you: migraine regresses to the mean. People start a prophylactic in a bad month, and the following three months are usually better whatever they take — which is precisely why the 28-day baseline is not optional and why the placebo arm improved at all Schoenen 1998.

What nobody has tested yet

Nobody has run the dose-response. Four hundred milligrams was chosen and it worked Schoenen 1998; 200 mg in children mostly did not Chen 2022. There has never been a trial randomizing adults to 25, 100 and 400 mg with monthly migraine days as the endpoint, which means the field does not know whether the effective dose is the absorbed ceiling or something well below it. Given a saturable transporter Zempleni 1996Yoshimatsu 2014, that is not a small gap — it is the difference between one tablet and eleven capsules.

Nobody has measured a flavin in a responder. Not one migraine trial reports plasma riboflavin, plasma FAD or the erythrocyte glutathione reductase activation coefficient in responders against non-responders. If the mechanism on this page is right, responders should be the people whose flavin status moved most, and a single additional blood tube in an existing trial would have answered it.

Nobody has stratified by mitochondrial genotype. The proposal that riboflavin response tracks with mitochondrial DNA haplogroup is twenty years old and has never been tested prospectively. A trial that genotyped at entry and reported response by haplogroup and by MTHFR 677 status would either turn this into the first genuinely personalized supplement recommendation on this site or retire the idea.

And nobody has tested split dosing. The entire absorption literature says the same daily total delivered in two or three portions gets more across the gut Zempleni 1996. Nobody has run 200 mg twice daily against 400 mg once daily for three months. That study is cheap, needs no new drug, and would change the label.

Riboflavin (B2) — its own safety story, not its category's

There is no upper intake level for riboflavin, and the reason is specific rather than reassuring. No tolerable upper limit has been set because no adverse effect has been established at any tested oral dose — and the mechanism that protects you is the same saturable transporter that limits the benefit Zempleni 1996Yoshimatsu 2014. You cannot easily poison yourself with a nutrient whose gut carrier stops accepting it. The absence of a limit is a statement about absorption, not a license.

What the yellow urine actually is. Filtered riboflavin, fluorescing. It is not detoxification, it is not the vitamin working, and its absence after a 400 mg dose would be the interesting finding — that would suggest the capsule did not dissolve.

The two real interactions, and one of them is usually stated backwards. Riboflavin is destroyed by light: this is why it is sold in opaque containers and why the milk industry moved away from clear glass. The tetracycline interaction that appears on this site's card and on most others is worth stating carefully, because the better-documented direction is the reverse of the common phrasing — tetracyclines and riboflavin taken together interfere with each other through metal-ion chelation in the gut, and the reported casualty is usually riboflavin absorption rather than antibiotic efficacy. The practical rule is identical either way: separate them by two hours. Note that a divalent-cation interaction is the same chemistry that makes tetracyclines incompatible with calcium, magnesium and iron, so a multivitamin is the bigger offender than a riboflavin capsule.

Two situations where the dose is a genuine question. Riboflavin and its photoproducts generate reactive oxygen under ultraviolet light — that is the entire basis of riboflavin/UVA corneal cross-linking as an ophthalmic procedure — so the theoretical concern about 400 mg a day plus heavy sun exposure is real chemistry, untested in people, and belongs in the labeled-extrapolation column rather than in a warning. And in pregnancy the migraine dose has not been studied; the nutritional dose obviously has.

What this bottle cannot do for you. It will not fix fatigue in someone who is not deficient, and treating persistent fatigue with a B vitamin instead of a ferritin, a thyroid panel and a sleep history is the expensive mistake this page would rather you did not make.

Sources read for this page

• Schoenen J, et al. Effectiveness of high-dose riboflavin in migraine prophylaxis. A randomized controlled trial.. Neurology 1998 · PMID 9484373
• Zempleni J, et al. Pharmacokinetics of orally and intravenously administered riboflavin in healthy humans.. American Journal of Clinical Nutrition 1996 · PMID 8604671
• Thompson DF, Saluja HS. Prophylaxis of migraine headaches with riboflavin: A systematic review. Journal of Clinical Pharmacy and Therapeutics 2017 · PMID 28485121
• Chen YS, et al. Effect of Vitamin B2 supplementation on migraine prophylaxis: a systematic review and meta-analysis. Nutritional Neuroscience 2022 · PMID 33779525
• Gaul C, et al. Improvement of migraine symptoms with a proprietary supplement containing riboflavin, magnesium and Q10: a randomized, placebo-controlled, double-blind, multicenter trial. Journal of Headache and Pain 2015 · PMID 25916335
• Yoshimatsu H, et al. Functional involvement of RFVT3/SLC52A3 in intestinal riboflavin absorption.. American Journal of Physiology: Gastrointestinal and Liver Physiology 2014 · PMID 24264046
• Office of Dietary Supplements, National Institutes of Health. Riboflavin 5'-Phosphate (Thorne) -- filed Supplement Facts panel: riboflavin (as riboflavin 5'-phosphate sodium) 36.5 mg per capsule, one capsule one to three times daily. NIH Dietary Supplement Label Database, DSLD ID 288632

How you would know if it worked

Riboflavin (B2) — safety & side effects

Bloodwork to run alongside Riboflavin (B2)

Baseline first, then again at 8–12 weeks.

The Full Micronutrient Screen panel covers these in one order — 11 markers, $363.60 with the discount applied.

Check results you already have → · All 103 markers A–Z

Riboflavin (B2) — frequently asked questions

What Riboflavin (B2) is used for

Riboflavin (B2) appears under 1 goal in the goal router.

Related Vitamins supplements

← Browse the full Supplement Vault

↑ Back to on this page