Does Melatonin Work the First Night?

Somebody swallowed a tablet at half past ten, lay there until one, and now wants to know whether they should give it another few nights. That is the question, and it deserves a straight answer rather than a protocol.

Here it is. There is nothing to wait for. Melatonin is not a sedative that builds up in you over a week, so the idea of a loading period is imported from a different kind of drug. But the first night cannot tell you much either — because the effect being argued over is so small that no single night could ever show it to you.

What melatonin is doing, and what it isn’t

Melatonin is a hormone your own body makes, mainly in the pineal gland, and its release is tied to the light-dark cycle: it climbs in darkness and light shuts it down. That is the whole biography. The evidence describes it as a chronobiotic — a timing signal — rather than a hypnotic acting on the brain’s arousal systems the way sedative-hypnotic drugs do.

Read that again with the bottle in your hand, because it quietly dismantles the promise on the front. A sedative works on you. A timing signal works on a schedule. If the mechanism is a nudge to a clock, then whether it does anything at all depends on whether your clock is the problem — and on when relative to your own rhythm you took it, since the direction and size of any phase shift depend on the timing of the dose against the individual’s circadian phase. Take it at the wrong point in your own cycle and the nudge does not simply fail; it can push the opposite way.

None of which is what “gentle, natural, restful” prepares you for.

The number the whole argument is about

The most-cited pooled analysis of melatonin for primary sleep disorders found a reduction in the time taken to fall asleep of 7.06 minutes against placebo — 95 per cent confidence interval 4.37 to 9.75 — and about 8.25 minutes more total sleep, with a small improvement in reported sleep quality.

That is a real finding. It is also seven minutes.

Sit with what seven minutes means for your first night. You cannot perceive it. You have no placebo arm lying in the other half of the bed. The night-to-night variation in how long any of us takes to fall asleep dwarfs it entirely, which means a first night that felt like a miracle and a first night that felt like nothing are both perfectly compatible with an average effect of seven minutes. The trial detected it by averaging across many people. You are one person, having one night, with no comparison.

This is the mismatch that generates almost every melatonin argument on the internet. The finding is genuine and tiny. The marketing is enormous. And the reader who says “it didn’t work for me” and the reader who swears by it may both be describing the same seven minutes from opposite sides.

Two independent guideline bodies looked at this evidence base and reached the unglamorous conclusion. The American Academy of Sleep Medicine’s 2017 pharmacologic guideline did not recommend melatonin for sleep-onset or sleep-maintenance insomnia in adults. The American College of Physicians’ 2016 guideline on chronic insomnia did not establish it as a recommended treatment either, finding the evidence insufficient. Neither of those is a finding that melatonin has been proven useless. “Insufficient evidence to recommend” is a statement about the state of the evidence — but it is also not the endorsement the aisle is trading on.

Why the first night lies to you in both directions

Sleep aids are almost never taken on a random Tuesday. They get taken after a run of bad nights, at the point where somebody has had enough. And runs of bad nights end — that is what a run is. Regression to the mean hands the credit for a normal recovery night to whatever was swallowed before it, and expectancy effects do the rest. Both are well documented in insomnia trials, which is precisely why those trials have placebo arms and your bedroom does not.

There is a second reason the first night is unreadable, and it has nothing to do with your biology. When 25 melatonin gummy products sold in the US were analysed, 22 of them — 88 per cent — were inaccurately labelled for melatonin content, with measured amounts ranging from a fraction of what the label declared to more than three times it. An earlier analysis of products sold in Canada found content varying substantially from the label in most of what it tested. So the honest description of your first night is: an unknown quantity of a timing signal, taken at an unmeasured point in your own circadian phase, judged against a memory of how you usually sleep.

Nothing in that sentence supports a verdict.

Where the evidence actually holds up

An audit that only says no is a position with footnotes attached, so here is the part that survives.

Systematic review evidence supports melatonin for reducing jet-lag symptoms after crossing multiple time zones. Clinical guidelines support strategically timed melatonin in delayed sleep-wake phase disorder — a recognised diagnosis in which habitual sleep and wake timing sit substantially later than conventional timing, not merely a preference for late nights.

Look at what those two have in common: both are clock problems. The evidence is strongest exactly where the mechanism says it should be, and thinnest where the marketing is loudest. That is not a coincidence, and it is the single most useful thing to take from this page. The question worth asking is not “did it work last night” but “is my problem a clock problem at all.”

If it is chronic insomnia rather than a displaced clock — sleep difficulty at least three nights a week for at least three months, with daytime consequences — then the guidelines point somewhere else entirely. Both the ACP and the AASM recommend cognitive behavioural therapy for insomnia as the first-line treatment, as strong recommendations, while the same bodies did not recommend melatonin for that condition. Standalone sleep-hygiene advice, the thing everyone gets handed first, is specifically not recommended as a treatment and performs worse than the structured therapy.

The catch is honest and worth printing: CBT-I is harder to get than a bottle. Access is limited relative to demand. A supplement is available at midnight and the therapy is not, and that gap is a large part of why the shelf exists.

Two things this book got wrong, and had to fix

This site’s whole claim is that the checking is shown rather than asserted, so two of the failures behind this article belong on the page.

The first draft of the book stated that the 2023 gummy analysis had detected serotonin in some of the products it tested. It had not. That study screened for serotonin and found none in its 25 products — the serotonin contamination belongs to the earlier Canadian analysis. The machine had merged two studies into one more alarming study. Corrected, and it is the more interesting version anyway: the label problem is bad enough without borrowing somebody else’s finding.

The second was a number. A 1995 trial was described as showing that 0.1 to 0.3 mg reproduced a normal night’s plasma melatonin. The check found the range overstated; what the trial supports is a dose as small as 0.3 mg. A tenth of a milligram is not a rounding error when the entire argument is about how small the physiological amount is compared with what is sold.

Neither correction changes the conclusion. Both change how much you should trust an unchecked confident sentence — including the ones on the bottle.

The line where this stops being a supplement question

The largest number in the melatonin literature is not about adults at all. US poison-centre calls involving paediatric melatonin ingestions rose 530 per cent between 2012 and 2021, across 260,435 reported ingestions. The mandatory caveat belongs right beside it: the large majority — 84.4 per cent — were asymptomatic, and rising call volume partly reflects how many homes now have the stuff in a kitchen drawer in a form that looks like a sweet. It is not evidence that a dose became more dangerous. It is evidence of how much of it is now within reach of a four-year-old.

And for adults, the first night is the wrong question entirely if the sleep problem has features that no supplement addresses. Loud habitual snoring, witnessed pauses in breathing, choking or gasping, waking unrefreshed after enough time in bed — those are recognised indications for clinical assessment, because obstructive sleep apnoea is diagnosed by a sleep study and is not ruled out by how you respond to a tablet. Falling asleep unintentionally in the day, particularly at the wheel, is a prompt-assessment matter, not a shopping one.

Long-term data on nightly use in healthy adults over years is limited — that is the honest state of it, in a market built on nightly use.

So: does it work the first night?

Partly, and less than you would need to know it did. If melatonin is going to do anything for you it does it from the start, because there is nothing to accumulate — but what it does, averaged over trials, is around seven minutes, taken from an unverified quantity at an unmeasured point in your own rhythm. One night cannot resolve that. Several nights barely can.

The better question is the one underneath: is your problem timing, or is it insomnia? For a clock knocked sideways by travel or sitting genuinely late, the evidence is real and this is the tool’s actual job. For months of lying awake, the guidelines are pointing at a therapy, not a bottle — and that is a conversation with your own GP rather than a purchase.

The full claim-by-claim audit is in Does Melatonin Actually Work?, and every one of the 85 claims in that book — including the ones that failed the check and what happened to them — is published at the melatonin verification ledger. If the pattern here is familiar, the same audit run on the supplement most often bought for the same reason is in Which magnesium is best for sleep?, and the question of what a “fact-checked” badge is actually worth is taken apart in What does “fact checked” actually mean?.

This page is machine-written and doctor-checked. The claims here were verified against primary sources, that check caught the first draft misattributing a laboratory finding and overstating a dose figure, and a practising doctor reviewed the result and is accountable for it. Corrections are posted publicly at jameswhitfieldauthor.com/corrections.