Melatonin is sold over the counter almost everywhere, which makes it easy to assume it's been thoroughly safety-tested. Three recent reviews suggest the picture is more incomplete than that — at every dose level.
At the doses most people actually take — the 0.15 mg to 12 mg range covering typical over-the-counter products — a systematic review of 37 randomized controlled trials found adverse events that were mostly mild. The most frequently reported were daytime sleepiness (1.66%), headache (0.74%), other sleep-related side effects (0.74%), dizziness (0.74%), and hypothermia (0.62%). More serious effects — agitation, fatigue, mood swings, nightmares, skin irritation, palpitations — were rare, and most side effects at any level resolved on their own within days or cleared up as soon as melatonin was stopped. The catch: most of these trials lasted 4 weeks or less, the longest ran 29 weeks, and only 8 of the 37 studies met good-quality standards for risk of bias — so this tells you about typical short-to-medium-term use, not what happens after months or years of nightly use.
The numbers on side effects at higher doses come from a separate review focused specifically on 10 mg or more in adults. Researchers pooled 79 randomized controlled trials, 3861 participants in total, spanning a wide range of medical conditions. The first finding here is itself a warning sign: 29 of those 79 studies (37%) didn't report on adverse events at all. Of the studies that did, only 4 met strict low-risk-of-bias criteria for the main analysis. Within that small, higher-quality subset, high-dose melatonin showed no significant increase in serious adverse events (rate ratio 0.88, 95% CI 0.52–1.50) or withdrawals due to side effects (rate ratio 0.93, 95% CI 0.24–3.56) — but it did show a significant increase in milder adverse events such as drowsiness, headache, and dizziness (rate ratio 1.40, 95% CI 1.15–1.69, p < 0.001).
A third, broader review helps place both safety pictures in context. Looking across 71 studies published between 2011 and 2024, melatonin's evidence base turns out to be strongest for sleep quality and certain neurodegenerative conditions, with some support for reducing preoperative anxiety and aiding recovery in athletes. For obesity, blood sugar control, and gut microbiome effects, the evidence is inconsistent — results depend heavily on diet and exercise factors that are hard to separate from melatonin's own effect. For cancer, psychiatric disorders, pregnancy, and pediatric use, evidence remains limited. For respiratory health, it's largely preclinical (lab and animal studies, not yet tested in people). Adverse effects noted across this broader literature include nightmares and grogginess, alongside the drowsiness, headache, and dizziness already flagged in the dose-focused review.
A note on what's circulating but not yet verifiable. A preliminary analysis presented at a cardiology conference in late 2025, using health records from over 130,000 adults, reported an association between a year or more of melatonin use and higher rates of heart failure diagnosis, heart failure hospitalization, and death from any cause. This has not been peer-reviewed or published in a journal, so it doesn't carry the same evidentiary weight as the reviews above — but it's a live area of investigation worth watching rather than dismissing.
What ties this together. Melatonin's reputation as a low-risk supplement is not unearned — the meta-analysis found no signal for serious harm at higher doses, and mild side effects like drowsiness are hardly a shock for a sleep aid. But "safe" and "well-studied long-term" are not the same claim. More than a third of high-dose trials didn't even track adverse events, and the areas where melatonin gets marketed most heavily today — metabolic health, immune support, general wellness — are exactly where the evidence is weakest.
Bottom line. At typical over-the-counter doses, melatonin's side effects are usually mild — daytime sleepiness, headache, dizziness, each affecting well under 2% of users in the trial evidence — and higher doses (≥10 mg) show no clear signal for serious harm either, aside from a real increase in those same mild effects. But the evidence at both dose levels comes almost entirely from short trials: more than a third of high-dose studies didn't even report side effects, and even the best low-dose evidence rarely tracked people for more than a few months. Its use for anything beyond sleep and select neurological conditions also rests on thinner evidence than its popularity suggests. Long-term daily use, especially at higher doses, is worth discussing with a doctor rather than assuming it's risk-free.
Medical note. This article is intended for general education purposes and should not be considered personal medical advice.