Melatonin Is Not a Sleeping Pill

The job description is wrong

Walk down the sleep aisle and melatonin sits with the sleeping pills. Same shelf, same promises on the box, same implied job: take this, fall asleep. That framing is so complete that when I named melatonin as one of the few supplements on my floor, the reasonable response was that I'd overrated a sleep aid.

I'd argue the opposite. Melatonin is underrated precisely because it keeps getting judged against a job it was never suited for. Filed as a sedative it looks like a weak product. Filed as what it actually is, a timing signal, it looks like one of the better-evidenced things in the aisle. Same molecule, same trials, opposite verdict, and the only thing that changed is what you asked it to do.

So let me do both readings, starting with the one that goes badly for me.

As a sleeping pill, it is mediocre, and I'll say so

The pooled evidence is not impressive. The largest meta-analysis of melatonin in primary sleep disorders, nineteen trials and 1,683 people, found it shortened the time to fall asleep by about seven minutes and added about eight minutes of total sleep, with a small improvement in overall sleep quality. Those effects are real, they held up rather than fading with continued use, and they are still seven minutes.

The American Academy of Sleep Medicine went further. In its 2017 guideline on drug treatment for chronic insomnia in adults, it suggests clinicians not use melatonin for either sleep onset or sleep maintenance insomnia. I want to be scrupulous about that, because it is easy to wave away and easy to overstate. It is a weak recommendation under the GRADE system, and the guideline says plainly that weak reflects the certainty of the evidence rather than proof that something does not work. Still, the sleep field looked at melatonin as a treatment for insomnia and declined to endorse it. If your only reason to take it is that you cannot fall asleep, the evidence is not on your side, and I am not going to pretend otherwise to protect my own position.

As a timing signal, it is a different molecule entirely

Here is the part that almost never makes it onto the box. The same academy, in its guideline on circadian rhythm sleep-wake disorders, positively endorses strategically timed melatonin. Not for insomnia. For delayed sleep-wake phase disorder, for blind adults whose clocks free-run past twenty-four hours, and for children and adolescents with irregular sleep-wake rhythm alongside neurological conditions. One body, two guidelines, one molecule, opposite conclusions, and the variable that separates them is the job, not the evidence quality.

Be careful with what that guideline does and does not say, because I am about to go further than it does. Those are diagnosed circadian disorders in defined clinical populations, and the guideline says nothing about a healthy adult taking melatonin as a general practice. It establishes what the molecule does. The decision to put it on a floor is mine, and I'll make that case on my own terms rather than borrowing theirs.

Travel is the cleanest demonstration. The Cochrane review on jet lag looked at ten randomized trials and concluded that melatonin is remarkably effective at preventing or reducing it, with a number needed to treat of two. Two. That is a better number than most things I coach. And the detail that matters most in that review is not the size of the effect, it is the direction: taken at the wrong time of day, melatonin made adaptation worse rather than better. A sedative taken at the wrong hour makes you sleepy at the wrong hour. Only a clock-setting signal can push you further out of sync. That behavior is the tell, and it tells you what the molecule actually does.

Your body runs on a clock that expects darkness at a certain hour, and melatonin is the message that the hour has arrived. Send it on schedule and the clock moves toward where you want it. Send it randomly and you are adding noise to a system that runs on regularity.

Why that puts it on my floor and not my flex pile

In my practice every pillar has a floor, meaning the small set of basics that carry most of the durable results, and a flex, meaning the optional edges you earn once the floor is solid. Most of what the optimization world sells is flex sold early. The Molecules floor is deliberately short: creatine, vitamin D with K2, magnesium, the electrolytes you genuinely miss, and a multivitamin as temporary coverage. Melatonin belongs on that list, and the reasoning is mine.

Circadian timing is upstream. Sleep quality sits downstream of when your body thinks night is, and almost everything else I coach sits downstream of sleep: recovery from training, appetite regulation, mood, the ability to hold a hard habit for more than a week. Most of the work on that clock is free, which is exactly why I like it. Get light in your eyes in the morning, get it out of them at night, keep your hours boring. Melatonin is the one purchasable thing that acts on that same system instead of sedating you on top of a badly timed life.

And I want to be direct about what I am not claiming. I am not telling you melatonin extends your life. No trial has shown that, and the antioxidant and cellular stories that get quoted in longevity circles are early, mostly preclinical, and nowhere near ready to hang a recommendation on. My floor claim is narrower and I think it is stronger: this molecule acts on the system that sets when you sleep, that system is load-bearing, and the cost of being wrong is small.

I'll also say what melatonin is not going to fix. For years my own sleep was fragmented while I worked late, traveled constantly, stayed out with teams, and drank more than I should have, and I expected the night to work anyway. No pill was going to solve that, and I've written about the whole thing in Sleep: The Foundation of the Mind. Build the floor first. The tools only work on top of it.

Honest Take

Melatonin and vitamin D are both Tier 1 work in my practice, and their evidence grades are not the same. That is not an inconsistency, it is the point of keeping the two scales separate. Tier is how foundational the work is. Grade is how strong the evidence is. Vitamin D has decades of human outcome data and a shortfall you can measure with a cheap blood test. Melatonin has good chronobiotic evidence, weak insomnia evidence, and essentially no long-term outcome data in healthy adults, because nobody has run that trial. Same tier, different grades, and anyone who tells you a molecule is "Tier 1 graded" has collapsed two scales that do different jobs.

The objections, taken seriously

It is a hormone. Yes it is, and this is the objection I respect most. The safety review literature says short-term use is safe in humans even at extreme doses, with mild effects like headache, dizziness, nausea and daytime sleepiness, and that randomized long-term treatment produces adverse effects comparable to placebo. That same review is explicit that long-term safety in children and adolescents needs further investigation, and that pregnant and breastfeeding women should not take it given the absence of human studies. So the honest position is that the short and medium-term safety record is good and the long-term record in healthy adults barely exists.

The 2025 heart failure headline. In November 2025 a study presented at the American Heart Association's Scientific Sessions reported that among 130,828 adults with chronic insomnia, those with a year or more of documented melatonin use had roughly a 90 percent higher rate of new heart failure and about double the rate of death from any cause over five years. That is a real finding and I am not going to hide from it. I am also going to tell you what it can and cannot establish. It is a conference abstract rather than a peer-reviewed paper, and the American Heart Association's own release says results like it stay preliminary until they publish as full manuscripts. I have not found a peer-reviewed publication of it. It is observational, and the association runs in the obvious direction of confounding: people who receive a year of prescribed melatonin for chronic insomnia are sicker sleepers than people who never needed it. And it has one specific defect the American Heart Association's own release names, which runs against the headline rather than for it. The database spans countries where melatonin requires a prescription and countries like the United States where you buy it off a shelf, and patient location was not available, so American over-the-counter users would have been sorted into the group counted as never having used it. That is the kind of misclassification that can manufacture a difference out of nothing. I am watching this one. If it publishes and survives review, I will change what I write here.

The bottle is unreliable. This is the caveat I think gets the least attention and deserves the most. When researchers analyzed thirty-one commercial melatonin supplements, the actual melatonin content ranged from 83 percent below the label claim to 478 percent above it. More than seven in ten missed their own label by more than ten percent. Lot-to-lot variation inside a single product ran as high as 465 percent. And serotonin, which is a controlled substance used to treat neurological conditions, turned up in eight of the samples. A separate analysis of twenty-five melatonin gummies sold in the United States found the labeling problem intact: actual melatonin ran from 74 percent to 347 percent of the labeled amount, 88 percent of the products were inaccurately labeled, and one contained no detectable melatonin at all. To be fair to that second study, it did not find serotonin in anything, so the contamination finding has not repeated. The labeling problem has.

That is also my answer to the question everyone asks, which is how much. I coach that inside a practice, with the individual in front of me and the physician they work with, and I deliberately do not publish a number here. Two reasons. The amount on a shelf is untethered from what your body actually makes, and after the analysis above you cannot be confident the bottle contains what it says. When the label is not reliable, a public number is false precision, and false precision is worse than no number.

Who should skip it. The circadian guideline recommends against melatonin for elderly patients with dementia, which is a genuine against rather than an absence of evidence. Children are their own conversation and not a casual one: pediatric melatonin ingestions reported to United States poison control centers rose 530 percent over ten years, and that report documents children who required mechanical ventilation and two who died. The jet lag review flagged case reports of harm in people with epilepsy and people taking warfarin. If you are pregnant or breastfeeding, if you are on anticoagulants or have a seizure disorder, this is a conversation with your physician and your coach before it is a purchase.

Where it sits with everything else

Melatonin lives in the Molecules pillar because it comes in a bottle, but the work it supports is Mind, where sleep is the foundation, and the payoff lands in Movement, because recovery from hard training is a nighttime process. It touches Meals too, since a late heavy dinner is its own signal to the same clock. That is the pattern across my practice: the molecule is rarely the intervention, it is a small assist on a system you are already running with free levers.

The honest bottom line

Melatonin is a weak sleeping pill and a good clock-setter, and almost everyone buys it for the first job. I take it, I recommend it to clients, and it sits on my Molecules floor because the system it acts on is upstream of most of the things people come to me for. The evidence behind the timing job is solid. The evidence behind the longevity claims is not there yet, and I am not going to borrow it. The product quality is a genuine problem and the dose question does not have an honest public answer.

This piece sits underneath Proven Before Promising, my Molecules overview, which is where the whole pillar starts and where the rule comes from that a molecule earns its place before anyone gets to promote it. If you have not read the rest of the floor, The Molecules Floor covers the short list this piece belongs to. If your actual problem is that your nights are not working, start with Sleep: The Foundation of the Mind instead, because that is the floor underneath this one. And if you want to know why I keep saying tier and grade are two different scales, that is How Hard, and How Sure.

Evidence

Highest evidence grade in this article: B · Strong

Each claim below is graded on its own; a high grade for one does not carry to the others.

• Strategically timed melatonin for circadian rhythm sleep-wake disorders (delayed sleep-wake phase disorder, non-24 in blind adults, irregular sleep-wake rhythm in children and adolescents with neurological conditions): C · Moderate (a guideline statement is graded by the evidence beneath it, and this one carries its endorsement at a second-tier degree of confidence, so it caps at C even though it rests on a systematic review of randomized trials)
• Melatonin for preventing or reducing jet lag: B · Strong (Cochrane review of ten randomized trials, number needed to treat of two)
• Melatonin for sleep onset latency, total sleep time and sleep quality in primary sleep disorders: C · Moderate (meta-analysis of randomized trials, but the effect sizes are small and on sleep parameters rather than a health outcome)
• The recommendation against melatonin for chronic insomnia in adults: C · Moderate (graded by the evidence beneath it, which the guideline itself rates as low certainty)
• The recommendation against melatonin in elderly patients with dementia: C · Moderate (guideline recommendation at a second-tier degree of confidence)
• Short and medium-term safety of melatonin in adults: C · Moderate (narrative safety review of randomized trials, not a quantitative meta-analysis)
• Long-term safety of melatonin in healthy adults, and any longevity or mortality benefit: not graded, because the trials do not exist. This is the honest gap in the piece.
• The reported association between long-term melatonin use and incident heart failure and all-cause mortality: not graded. It is an unpublished conference abstract from observational records, which does not meet the bar for a grade on this site, and its own release names a misclassification problem that runs against the finding.
• Melatonin content variability against label claim: C · Moderate (direct chemical analysis, which is a measurement rather than an inference, but each study is a single laboratory analyzing a few dozen products; two independent analyses agree on the labeling problem)
• Serotonin contamination of melatonin supplements: D · Emerging (found in one analysis, explicitly not detected in the second, so the finding has not replicated)
• Rising pediatric melatonin ingestions and associated harm: C · Moderate (national poison control surveillance data, which is observational)
• Antioxidant, mitochondrial and longevity mechanisms for melatonin: E · Frontier (preclinical and mechanistic, and deliberately not used to support anything in this piece)
• Placing melatonin on the Molecules floor as Tier 1 work: this is my coaching position, built on the circadian evidence above plus my own practice with clients and the physicians they work with. It is not a trial finding and I have not presented it as one.

Sources

• Auger RR, Burgess HJ, Emens JS, Deriy LV, Thomas SM, Sharkey KM. Clinical practice guideline for the treatment of intrinsic circadian rhythm sleep-wake disorders: advanced sleep-wake phase disorder (ASWPD), delayed sleep-wake phase disorder (DSWPD), non-24-hour sleep-wake rhythm disorder (N24SWD), and irregular sleep-wake rhythm disorder (ISWRD). An update for 2015: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2015;11(10):1199-1236. https://doi.org/10.5664/jcsm.5100
• Sateia MJ, Buysse DJ, Krystal AD, Neubauer DN, Heald JL. Clinical practice guideline for the pharmacologic treatment of chronic insomnia in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2017;13(2):307-349. https://doi.org/10.5664/jcsm.6470
• Herxheimer A, Petrie KJ. Melatonin for the prevention and treatment of jet lag. Cochrane Database Syst Rev. 2002;(2):CD001520. https://doi.org/10.1002/14651858.CD001520
• Ferracioli-Oda E, Qawasmi A, Bloch MH. Meta-analysis: melatonin for the treatment of primary sleep disorders. PLoS One. 2013;8(5):e63773. https://doi.org/10.1371/journal.pone.0063773
• Andersen LPH, Gögenur I, Rosenberg J, Reiter RJ. The safety of melatonin in humans. Clin Drug Investig. 2016;36(3):169-175. https://doi.org/10.1007/s40261-015-0368-5
• Erland LA, Saxena PK. Melatonin natural health products and supplements: presence of serotonin and significant variability of melatonin content. J Clin Sleep Med. 2017;13(2):275-281. https://doi.org/10.5664/jcsm.6462
• Cohen PA, Avula B, Wang YH, Katragunta K, Khan I. Quantity of melatonin and CBD in melatonin gummies sold in the US. JAMA. 2023;329(16):1401-1402. https://doi.org/10.1001/jama.2023.2296
• Lelak K, Vohra V, Neuman MI, Toce MS, Sethuraman U. Pediatric melatonin ingestions, United States, 2012-2021. MMWR Morb Mortal Wkly Rep. 2022;71(22):725-729. https://doi.org/10.15585/mmwr.mm7122a1
• Nnadi E, Masara M, Offor R, Unal S, Rebah R, Atere M, Nigussie B, Graham-Hill S. Abstract 4371606: effect of long-term melatonin supplementation on incidence of heart failure in patients with insomnia. Circulation. 2025;152(Suppl 3):4371606. Presented at American Heart Association Scientific Sessions 2025. Conference abstract, not peer reviewed. https://doi.org/10.1161/circ.152.suppl_3.4371606
• American Heart Association newsroom. Long-term use of melatonin supplements to support sleep may have negative health effects. November 3, 2025. https://newsroom.heart.org/news/long-term-use-of-melatonin-supplements-to-support-sleep-may-have-negative-health-effects