The dose of melatonin your body responds to and the dose sold in the sleep aisle are not the same number. We wanted to know how far apart they actually are, so we counted.
The National Institutes of Health keeps a public database of supplement labels called the DSLD. We pulled 600 melatonin products currently on the market and read the supplement facts panel on each one. 588 stated a dose we could read in milligrams. Here is what the shelf looks like.
The median melatonin dose is 3 mg
588 labels · NIH database
Where the melatonin doses actually sit
Every melatonin product we could read a dose from, grouped by milligrams per serving.
The gold row is the one the research used. A 0.3 mg dose restores blood melatonin to the normal nocturnal range. Eleven products out of 588 sell at or below 0.5 mg — the median on the shelf is 3 mg, ten times higher.
The doses ran from 0.1 mg to 60 mg. That is a six-hundred-fold spread inside a single category, all of it legal, all of it sitting on the same shelf.
Some quick landmarks from the same 588 labels:
- 77.7% are at 3 mg or above.
- 45.1% are at 5 mg or above.
- 18.2% are at 10 mg or above.
- 1.9% — eleven products out of 588 — are at 0.5 mg or below.
Why 0.3 mg is the number worth knowing
Melatonin is unusual among supplement ingredients: it isn’t a nutrient or a plant extract, it’s a hormone your own brain releases as light fades. So there is a meaningful question you can ask about it that you can’t ask about most ingredients — what dose reproduces what the body does on its own?
That question has been studied directly. In a trial at MIT, healthy young adults were given 0.3 mg or 1.0 mg of melatonin a few hours before their usual bedtime. The 0.3 mg dose raised blood melatonin into the normal nighttime range — about 113 pg/ml — and shortened the time it took to fall asleep, with no measurable hangover the next morning on either mood or performance tests (PMID 8843534).
A later trial from the same group is the more interesting one for anyone who wakes in the middle of the night. Adults over 50 with reduced sleep efficiency were given placebo, 0.1 mg, 0.3 mg and 3.0 mg in random order, each for a week, with sleep measured by polysomnography (PMID 11600532). The physiological dose — 0.3 mg — restored sleep efficiency and, in the authors’ words, acted principally in the middle third of the night. It also brought blood melatonin back to a normal nighttime level.
The honest part: 3 mg worked too
We’re not going to trim the result to fit an argument. In that same trial the 3.0 mg dose also improved sleep. Higher was not useless.
What higher did do was two other things. It induced hypothermia — a drop in body temperature. And it left melatonin elevated in the blood into the daylight hours, which is the mechanism behind the morning fog we wrote about separately. You got the sleep benefit and a hormone signal that was still running when you needed it switched off.
So the finding isn’t “melatonin doesn’t work” and it isn’t “high doses are dangerous.” It’s narrower and, we’d argue, more useful: past the physiological dose, you stop buying more effect and start buying more duration.
What the label numbers are actually based on
Here is the part that bothers us. If you set out to pick a melatonin dose from the research, you would land somewhere near 0.3 mg — that’s where the plasma levels land in the normal range and where the mid-night effect was demonstrated. Almost nobody sells that. It’s under 2% of the labels we read.
Instead the shelf clusters at 3, 5, and 10 mg — numbers that are round, easy to print, and easy to make sound stronger than the bottle beside them. A 10 mg product isn’t ten times the product of a 1 mg one. It’s the same hormone, given at a level your body has no template for, staying in circulation longer.
This is the same pattern we found when we read 336 ashwagandha labels, and it’s the reason we keep counting instead of assuming. In that category the doses were mostly fine and the problem was that 71% of labels couldn’t be compared to any study at all. Here the doses are perfectly clear and the problem is that the number on the front was chosen by the marketing convention of the category rather than by the research.
Two different failures. Same underlying habit: the number on the label is a claim about strength, not a claim about evidence.
Why our bottle doesn’t contain any
We make an evening supplement, and it has no melatonin in it. Given everything above, the obvious question is why not — 0.3 mg is cheap, well-studied, and works in exactly the part of the night we built the product around.
We thought about it seriously, and it’s on our list of ingredients we rejected along with eleven others. The reason we said no isn’t that the research is weak. It’s that the category has trained people on 3 to 10 mg, that’s the experience they associate with the word melatonin, and it’s the specific experience — waking up foggy — that most of our customers are trying to get away from. Putting a melatonin line on our label would have undone the reason they came, no matter how small the number after it.
Our formula works on the stress side of the night instead — the cortisol rise that collides with lightening sleep around 3 a.m. — which is a different lever entirely.
And to be plain about it: if you want to try 0.3 mg of melatonin on its own, we think that’s a reasonable thing to do. It costs a few dollars, the evidence behind it is real, and it isn’t ours.
How we counted, so you can check us
We searched the DSLD for “melatonin”, filtered to products currently on the market, and read 600 labels on 29 August 2026. The database held 1,804 matching records, so this is a sample, not the whole shelf.
For each label we took the melatonin line from the supplement facts panel and converted micrograms to milligrams. Twelve labels stated no readable dose and were excluded. One label — an AvKare unit-dose product titled “Maximum Strength Melatonin 5 mg” — carried 50 mg in its panel, which is the package count of 50 tablets sitting in the dose field; we corrected it to 5 mg and note it here rather than quietly dropping it. Correcting it changed no conclusion in this article. We checked every dose above 15 mg and every dose below 1 mg against the product name by hand; apart from that one record, they agree.
The database is free and needs no account. If you repeat this and get different numbers, we’d like to know.
†These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease. This article is educational and is not medical advice. Melatonin is a hormone; if you are pregnant or nursing, under 18, taking medication, or managing a health condition, talk to your healthcare provider before using it.
We ran the same count across the wider category. In 600 sleep formulas, the pattern flips: melatonin products are simple and plainly dosed, while multi-ingredient sleep formulas hide 38.6% of their doses inside blends.
One more thing we later found: melatonin is the only common sleep ingredient whose dose does not shrink inside multi-ingredient formulas — the expensive actives dilute, the cheap hormone keeps its number.
Credit where due: in our warnings census, melatonin turned out to be the best-warned ingredient on the shelf — 81% of labels warn about drowsy driving.