This is one of the most common questions people ask before buying a melatonin product. We sell two and have never answered it — so here it is, including the parts arguing against ours.
It arrives as one question when it is really four: will I become physically dependent, will my sleep worsen if I stop, will it stop working, and will I end up unable to sleep without a pill. Three are better supported than the fourth.
Start with what melatonin is
Melatonin is a hormone your pineal gland already makes on a daily schedule. That matters, because the drugs people fear — benzodiazepines and Z-drugs — amplify inhibition in the brain, and their tolerance and withdrawal syndromes come from the brain adapting to that. Melatonin is not doing that job. A 2026 review in Clocks & Sleep puts it bluntly: melatonin "acts primarily as a chronobiotic rather than a hypnotic," with effects depending more on when you take it than how much.
Regulators read it differently by region. Melatonin is regulated as a medicine in the UK and the EU and sold as an over-the-counter dietary supplement in the US and Canada. In Ireland it is prescription-only and not permitted in food supplements at all. Across the EU, preparations of 2mg and above are medicinal products under European Medicines Agency oversight; lower doses are often sold as food supplements.
Physical dependence: what has been studied, and what has not
The best long-term data comes from prolonged-release melatonin, the prescription 2mg formulation used in Europe. In an open-label study of 244 adults with primary insomnia, participants took it nightly for 6 or 12 months, followed by a two-week withdrawal period and a structured symptom checklist. The authors reported no evidence of tolerance, and discontinuation was not associated with rebound insomnia or withdrawal symptoms — they observed residual benefit instead. A 6-month randomised placebo-controlled trial in 791 adults aged 18 to 80 agreed: effects were "maintained or enhanced over the 6-month period with no signs of tolerance."
Now the limits, because this is where a company selling melatonin would normally stop writing. The 244-person study was open-label, by authors affiliated with the manufacturer of the product tested. Both studies used 2mg prolonged-release melatonin, built to mimic the body's own overnight curve. Neither tells you anything about 10mg of immediate-release melatonin inside a botanical blend, which is what most supplements — including one of ours — deliver. A 2022 review in Expert Opinion on Drug Safety that hunted for long-term prospective trials with systematic adverse-event monitoring concluded that while event rates are low, "the scarcity of data from double-blind randomized placebo-controlled trials should caution against complacency." No dependence syndrome has been described. The evidence is thinner than the confidence with which it gets cited — and note which way the research runs: melatonin has been studied as a way to help people come off benzodiazepines, not the reverse.
Rebound insomnia is a different question
Rebound means your sleep comes back worse than before, specifically because you stopped. It is documented with some hypnotics and is not dependence. The melatonin data is consistent: a three-week randomised placebo-controlled trial in 170 insomnia patients aged 55 and over specifically evaluated rebound insomnia and withdrawal effects and found neither, and the 6–12 month study found the same after a year. The 2026 review summarises it as: "Across adult and pediatric studies, prolonged-release melatonin, even when used long-term, has not been associated with physiological withdrawal or significant symptom rebound" — adding, in a caveat worth repeating rather than burying, that this is "context-specific and not as justification for indiscriminate dose escalation."
There is a trap here. If you stop and sleep badly, the likeliest explanation is not rebound but the untreated problem returning — and from the inside those feel identical. What actually helps with 3am waking is a different toolkit.
Does it stop working?
Two halves. First: the trial evidence does not show tolerance developing out to 6–12 months, and both long-term studies looked for it. We also looked for human evidence that melatonin receptors downregulate under sustained exposure — the mechanism you would expect if tolerance were real — and found none we would stand behind.
Second: past roughly a year, nobody knows. Most melatonin trials run for weeks. The 2024 dose-response meta-analysis behind the best estimate of melatonin's effective dose pooled 26 randomised double-blind trials published between 1987 and 2020 — substantial work, almost none of it long. If you have taken melatonin nightly for three years, you are outside the evidence.
There is a more mundane explanation for "it stopped working," probably right more often than tolerance. Melatonin is a clock signal, not a sedative, so taking it 30 minutes before bed uses it as the wrong thing. That same analysis found the interval between dosing and the sleep episode was a significant predictor of sleep-onset improvement (β = −0.16, p = 0.023), pointing toward roughly three hours before target bedtime. The Clocks & Sleep review argues mistimed dosing produces "avoidable treatment failure" and "inaccurate labeling of nonresponse." Our breakdown of the dose-response evidence and our guide to resetting a shifted sleep clock work through the timing logic.
Does it shut down your own production?
This is asserted constantly, and the evidence is narrower than the claim. A 1997 study gave 0.5mg of melatonin or placebo at bedtime to 21 night-shift workers for seven days, then measured their endogenous profiles; the amplitude of their own secretion was unchanged. The same paper reports one blind participant given 50mg nightly for 37 days, with no change. The authors concluded that circulating melatonin "can shift the phase, but does not alter the amplitude, of pineal melatonin secretion." Separately, the 6–12 month study measured nocturnal urinary 6-sulfatoxymelatonin, the marker of overnight output, after up to a year of nightly 2mg: no suppression.
The evidence points one way, and there is very little of it. Twenty-one shift workers, one blind participant and one open-label cohort is not proof that nothing happens over years. No study offers certainty that your own production is untouched after years of 10mg doses.
The reliance nobody writes about
The fourth question has the least research and the most relevance. Psychological reliance — needing the ritual rather than the molecule — is not dependence, and there is essentially no melatonin-specific trial literature on it. We are not going to invent one. But it is the honest answer for many, and the mechanism is not mysterious: if the pill is the signal that the day is over, the conditioning is doing work the milligrams are not. Our piece on the placebo effect in supplements covers why that is neither fake nor trivial, and suggests a free test: keep the routine, drop the capsule, watch two weeks. On whether supplements need cycling, melatonin's answer is simply: take it when it has a job.
Dose matters more than duration
Worry about the number on the label rather than the number of nights. The dose-response meta-analysis found melatonin's effect on sleep onset and total sleep time rose with dose and peaked at 4mg/day; the 2026 review reads that literature the same way, putting the plateau at around 4mg in adults, with higher doses offering "little additional benefit." Melatonin follows first-order kinetics: a bigger dose gives proportionally higher blood levels, not proportionally more sleep. And the adverse-event signal is real: a 2022 review pooling 79 randomised trials of 10mg-plus doses in 3,861 adults found increased risk of mild events like drowsiness, headache and dizziness (rate ratio 1.40, 95% CI 1.15 to 1.69, p < 0.001), with no detectable increase in serious events (0.88, 0.52 to 1.50, p = 0.64). Only four of the 79 met the low risk-of-bias threshold, and 29 never mentioned adverse events.
That lands on our own shelf. Sleep Formula contains 2mg of melatonin, below the 4mg peak. Sleep Support contains 10mg, above it — not dangerous on the safety data, but not buying extra benefit either, and squarely in the band where that mild-adverse-event signal was measured. Both numbers are on the labels, and our head-to-head comparison lays out the trade-off. Never take both on the same night: that is 12mg in one dose, three times the studied peak, with no research behind the combination. If your conclusion is less melatonin or none, we would rather you got there — the non-melatonin options with real trial support are here for that reason.
Children and adolescents
Melatonin use in anyone under 18 is a clinical decision and we are not going to give guidance on it. Our labels say under-18s should consult a physician, and that is our whole position. The American Academy of Sleep Medicine's advisory says parents should talk to a health care professional before giving melatonin or any supplement to a child, and notes that in the US melatonin "is not under FDA oversight like other over-the-counter (OTC) or prescription medications." A 2023 systematic review with GRADE assessment found, across 22 randomised studies in 1,350 children and adolescents, no association with serious adverse events but an increased rate of non-serious ones (relative risk 1.56, 95% CI 1.01 to 2.43, 17 studies, I² = 47%) — and on pubertal development, only four observational studies totalling 105 patients. Its authors called for "caution against complacent use." The paediatric evidence is strongest in the melatonin research in autism — but that is a conversation for a clinician.
Bottom line
On the available evidence: melatonin does not produce the dependence syndrome people fear, no rebound insomnia has appeared in the trials that looked for it, tolerance has not appeared out to 6–12 months, and the narrow evidence on endogenous production points to no suppression. Every one of those findings is thinner than it sounds — mostly 2mg prolonged-release, mostly under a year, some open-label and manufacturer-affiliated.
The reliance most likely to be real for you is the psychological kind, and the test costs nothing. And if the problem is chronic insomnia rather than a run of bad nights, no dose of melatonin is the answer — CBT-I has substantially better evidence than any supplement, including both of ours.
This article is for informational purposes only and has not been evaluated by the FDA. It is not intended to diagnose, treat, cure, or prevent any disease. Speak with a healthcare provider before starting or stopping any supplement, especially if you take medication or have an existing health condition. Melatonin use in children and adolescents should be discussed with a physician.
Sources: Cruz-Sanabria F, Bruno S, Crippa A, Frumento P, Scarselli M, Skene DJ, Faraguna U. "Optimizing the Time and Dose of Melatonin as a Sleep-Promoting Drug: A Systematic Review of Randomized Controlled Trials and Dose-Response Meta-Analysis." Journal of Pineal Research, 2024;76:e12985. doi:10.1111/jpi.12985. — Lemoine P, Garfinkel D, Laudon M, Nir T, Zisapel N. "Prolonged-release melatonin for insomnia — an open-label long-term study of efficacy, safety, and withdrawal." Therapeutics and Clinical Risk Management, 2011. doi:10.2147/TCRM.S23036. — Wade AG, Ford I, Crawford G, McConnachie A, Nir T, Laudon M, Zisapel N. "Nightly treatment of primary insomnia with prolonged release melatonin for 6 months: a randomized placebo controlled trial on age and endogenous melatonin as predictors of efficacy and safety." BMC Medicine, 2010;8:51. doi:10.1186/1741-7015-8-51. — Lemoine P, Nir T, Laudon M, Zisapel N. "Prolonged-release melatonin improves sleep quality and morning alertness in insomnia patients aged 55 years and older and has no withdrawal effects." Journal of Sleep Research, 2007;16:372–380. doi:10.1111/j.1365-2869.2007.00613.x. — Matsumoto M, Sack RL, Blood ML, Lewy AJ. "The amplitude of endogenous melatonin production is not affected by melatonin treatment in humans." Journal of Pineal Research, 1997. doi:10.1111/j.1600-079X.1997.tb00301.x. — Menczel Schrire Z, Phillips CL, Chapman JL, Duffy SL, Wong G, D'Rozario AL, Comas M, Raisin I, Saini B, Gordon CJ, McKinnon AC, Naismith SL, Marshall NS, Grunstein RR, Hoyos CM. "Safety of higher doses of melatonin in adults: A systematic review and meta-analysis." Journal of Pineal Research, 2022. doi:10.1111/jpi.12782. — Besag FMC, Vasey MJ. "Adverse events in long-term studies of exogenous melatonin." Expert Opinion on Drug Safety, 2022;21. doi:10.1080/14740338.2022.2160444. — Kalkanis A, Karkala A, Pataka A. "Melatonin in Clinical Practice: Grey Zones Between Chronobiology, Insomnia and Consumer Supplementation." Clocks & Sleep, 2026;8(3):38. doi:10.3390/clockssleep8030038. — Morera-Fumero AL, Fernandez-Lopez L, Abreu-Gonzalez P. "Melatonin and melatonin agonists as treatments for benzodiazepines and hypnotics withdrawal in patients with primary insomnia. A systematic review." Drug and Alcohol Dependence, 2020;212:107994. doi:10.1016/j.drugalcdep.2020.107994. — Händel MN, Andersen HK, Ussing A, Virring A, Jennum P, Debes NM, Laursen T, Baandrup L, Gade C, Dettmann J, Holm J, Krogh C, Birkefoss K, Tarp S, Bliddal M, Edemann-Callesen H. "The short-term and long-term adverse effects of melatonin treatment in children and adolescents: a systematic review and GRADE assessment." EClinicalMedicine, 2023;61:102083. doi:10.1016/j.eclinm.2023.102083. — American Academy of Sleep Medicine, "Health Advisory: Melatonin Use in Children and Adolescents," 2022 (also Journal of Clinical Sleep Medicine, doi:10.5664/jcsm.10332). — Food Safety Authority of Ireland, food alert on melatonin as a prescription-only medicinal product not permitted in Irish food supplements.