Melatonin is a timing signal, not a sleeping pill. At physiological doses it tells the clock that night has started. At supplement-aisle doses it floods the system for hours longer than your own pineal gland ever would.
Your pineal gland begins releasing melatonin a couple of hours before your habitual bedtime, as light fades. It peaks in the middle of the night and is close to undetectable by morning. The signal it sends is about time rather than sleepiness: it tells the suprachiasmatic nucleus, and every tissue that listens to it, that the biological night has begun.
Richard Wurtman's group at MIT spent the 1990s working out how much melatonin it takes to reproduce that signal from outside. The answer was about 0.3 milligrams, which raised blood levels into the normal night-time range. Irina Zhdanova and Wurtman's 2001 trial in older adults with insomnia found that 0.3mg improved sleep efficiency, while 3mg was no better and left blood levels elevated into the following day.
Most melatonin sold in the US comes in 3, 5 or 10mg tablets. In the UK it is prescription-only, which is a useful indication of how differently regulators view the same molecule.
A clock signal sold
as a sedative
Melatonin acts mainly on two receptors, MT1 and MT2. MT1 activity is associated with dampening the alerting signal from the circadian clock in the evening. MT2 is the one most involved in shifting the timing of the clock itself. That is why melatonin's best-established use is moving the clock rather than knocking you out.
Timing determines the direction of the shift. Taken in the early evening, several hours before your own melatonin rises, it advances the clock and makes you sleepy earlier. Taken in the morning, it delays it. Taken at bedtime by someone whose clock is already normal, it mostly overlaps with melatonin the body was about to make anyway.
The carousel says chronic high-dose use desensitises MT1 and MT2 receptors, so the same dose does progressively less. Receptor desensitisation after prolonged melatonin exposure is well documented in cell and animal studies. It has not been demonstrated as a clinical tolerance effect in humans. The mechanism is plausible. The human consequence is unproven.
What is clearer is the pharmacokinetics. Melatonin has a short half-life, under an hour, but a large dose means levels can stay above the normal night-time range for many hours, sometimes into the morning. Zhdanova's group measured exactly that at 3mg. Residual melatonin at seven in the morning is a reasonable explanation for the grogginess some users report, and it is also a signal arriving at the wrong time of day.
Does it help you sleep?
A little.
Amnon Brzezinski, Wurtman and colleagues pooled seventeen studies in Sleep Medicine Reviews in 2005. Melatonin reduced the time taken to fall asleep by about four minutes, increased total sleep by about thirteen minutes, and improved sleep efficiency by around two percentage points. Eduardo Ferracioli-Oda's 2013 meta-analysis in PLoS One, covering nineteen trials and nearly sixteen hundred people with primary sleep disorders, found a reduction of about seven minutes in sleep onset.
Statistically real and clinically small. Neither analysis found that higher doses produced larger benefits, which is the finding that matters most for anyone holding a ten milligram tablet. The American Academy of Sleep Medicine's 2017 guideline advises against melatonin for chronic insomnia in adults because the evidence of meaningful benefit is weak.
Then there is what is actually in the bottle. Lauren Erland and Praveen Saxena tested thirty one melatonin supplements in Canada for the Journal of Clinical Sleep Medicine in 2017. Actual content ranged from 83 percent below the label to 478 percent above it, varied by up to 465 percent between lots of the same product, and a quarter of the products contained serotonin. In the US, paediatric melatonin ingestions reported to poison centres rose by 530 percent between 2012 and 2021, mostly from gummies.
The carousel says high-dose melatonin causes REM suppression, lower HRV and disrupted sleep architecture. Melatonin is not known to suppress REM; some studies suggest it may slightly increase it, and the HRV link is not established. It also says a significant portion of chronic users report dependency. Melatonin does not produce dependence in the pharmacological sense. Both are deleted.
“Melatonin is a genuinely useful molecule being used for the wrong job at the wrong dose. For jet lag and delayed sleep phase it can work very well. As a nightly ten milligram sleeping pill for someone whose clock is fine, it is mostly an expensive way to feel groggy at breakfast.”
Five things, in order
of how much they matter
Fix the light before you reach for the bottle
Light is what your own melatonin responds to. Bright evening light delays and suppresses it; morning daylight sets the clock so it rises on time. Getting the light environment right does the job exogenous melatonin is trying to imitate, without the dosing problems.
Make the last two hours dim
Ordinary room lighting before bed suppresses melatonin and shortens its night-time duration. Joshua Gooley and colleagues at Harvard showed in 2011 that typical indoor light before bed suppressed melatonin onset in most participants. Dimming the house costs nothing and works with the signal your body already makes.
Use melatonin for timing problems, not as a nightly default
The Cochrane review of jet lag found melatonin effective for flights crossing five or more time zones, particularly eastward. It is also used clinically for delayed sleep phase. These are clock problems, which is what melatonin addresses. Nightly use by someone whose clock is already on time has the weakest case.
If you use it, use a low dose
0.3 to 0.5mg matches the physiological range and was as effective as larger doses in the MIT work. Higher doses did not perform better in the meta-analyses and leave more melatonin circulating into the morning. In the UK this is a conversation with a GP; elsewhere, pick a product with third-party testing given the content variability.
Take chronic insomnia to CBT-I
If you have needed something to sleep most nights for three months, melatonin is not the treatment the guidelines recommend. Cognitive behavioural therapy for insomnia is first-line in both US and European guidance, and it treats the problem rather than the symptom.
A 30-second decision rule
- Are you crossing time zones, or is your sleep consistently shifted late but normal once it starts?
- Are you taking more than 1mg?
- Have you been taking it most nights for more than a month to fall asleep?
Your body makes about a third of a milligram a night. The tablet in your drawer is thirty times that, and the extra isn't doing anything for your sleep.
Send this to your melatonin addict friend.
- Brzezinski, A. et al. (2005). Effects of exogenous melatonin on sleep: a meta-analysis. Sleep Medicine Reviews, 9(1).
- Zhdanova, I.V. et al. (2001). Melatonin treatment for age-related insomnia. Journal of Clinical Endocrinology & Metabolism, 86(10).
- Erland, L.A.E. & Saxena, P.K. (2017). Melatonin natural health products and supplements. Journal of Clinical Sleep Medicine, 13(2).
- Ferracioli-Oda, E. et al. (2013). Meta-analysis: melatonin for the treatment of primary sleep disorders. PLoS One, 8(5).
- Herxheimer, A. & Petrie, K.J. (2002). Melatonin for the prevention and treatment of jet lag. Cochrane Database of Systematic Reviews.
- Lelak, K. et al. (2022). Pediatric melatonin ingestions, United States, 2012-2021. MMWR, 71(22).
