Melatonin is safe to take every night

In adults, most trials that gave melatonin every night lasted a few weeks, and across them it was well tolerated: side effects were uncommon and mostly mild, chiefly daytime sleepiness, headache and dizziness. The largest long trial gave 791 adults with insomnia a 2 mg slow-release tablet or a placebo nightly for up to six months, and adverse events, including serious ones, were no more common on melatonin than on placebo. Only a few small randomized trials ran longer. In a one-year trial of 81 postmenopausal women, adverse events and falls were as frequent on placebo as on melatonin. In a trial of 189 care home residents, most with dementia, who took melatonin for 15 months on average, side-effect ratings did not rise, but mood was rated worse and withdrawn behaviour was more common than on placebo. For use over many years the evidence is observational, which cannot separate the effect of melatonin from the reasons people take it. Three large US cohorts followed for up to 23 years found no link between regular use and heart disease or type 2 diabetes. An analysis presented at a 2025 cardiology conference, for which no peer-reviewed report was found in October 2026, reported more heart failure among adults with insomnia who had used melatonin long term, and a 2026 health-record study found more dementia diagnoses among older adults prescribed melatonin than among those prescribed other sleep medicines, a result its authors caution against reading as an effect of melatonin.

The largest gaps concern children and a few specific groups. In children, trials lasting weeks found no rise in serious side effects but more mild ones than on placebo, and only four small follow-up studies have looked at puberty after years of use: three found little or no effect, one found a tendency toward later puberty after more than seven years of use, and reviewers rated that evidence very low certainty. The US National Center for Complementary and Integrative Health states that information on long-term safety is lacking, that people with epilepsy or taking blood thinners need medical supervision, that research in pregnancy and breastfeeding is lacking, and that melatonin can stay active longer in older people and cause next-day drowsiness. What is in the bottle can also differ from the label: 22 of 25 US gummy products tested in 2023 did not match their labels, with amounts up to about three and a half times the stated dose.

The broadest look at adults, a 2019 review of 37 placebo-controlled trials at doses from 0.15 to 12 mg a day, found daytime sleepiness in 1.66 percent of participants and headache and dizziness in 0.74 percent each, with a few more significant reports such as agitation, mood swings, nightmares and palpitations; most effects faded on their own or when melatonin was stopped. Participants were followed for up to 29 weeks and most trials ran 4 weeks or less, which the authors say limits conclusions about continuous use over longer periods. A 2022 review by two of the same authors, limited to studies lasting six months or longer, reports that possible adverse events were infrequent and few were clinically significant, and that blinded, placebo-controlled data on long-term use are scarce. In the six-month trial, 2.8 percent on melatonin and 5.1 percent on placebo had a serious adverse event during the extension; one case of palpitations, in a woman with a history of them, was judged possibly related. That trial was funded by Neurim Pharmaceuticals, which makes the tablet, and its founder and two of its employees were among the authors.

The two longer trials cited here were small and in specific groups. The one-year trial gave 1 or 3 mg of melatonin or a placebo nightly to postmenopausal women with low bone density; it recorded 36 adverse events and 12 serious ones, at equal frequency in both groups, two falls in each group, and no difference in balance, muscle strength or quality of life, and it was supported by Danish research funds and foundations. The care home trial ran in the Netherlands from 1999 to 2004 in residents with an average age of 86, who took 2.5 mg each evening for up to three and a half years while their facilities were assigned to bright or dim daytime lighting. The withdrawn behaviour was not seen when melatonin was combined with bright light, and the authors recommended melatonin only together with light for that group. It was funded mainly by Dutch public research bodies and foundations, with three equipment companies supplying material at reduced cost.

The 23-year cohort study covered 159,072 US health professionals and recorded melatonin only as a yes or no answer to regular use, with no information on dose, frequency or how long people took it. The 2025 heart failure analysis drew on electronic health records of adults with insomnia. A 2026 peer-reviewed article describing it notes that people who take melatonin differ from those who do not, particularly in sleep problems and depression, and its authors' own analysis of US national survey data found no significant association between recent melatonin use and heart disease or heart failure. The 2026 dementia study matched about 190,000 adults aged 50 and over with sleep disorders who were newly prescribed melatonin with the same number newly prescribed a benzodiazepine, and dementia was diagnosed about twice as often in the melatonin group over a few years of records; its authors write that disturbed sleep can be an early sign of the diseases being counted and that the people given melatonin may have differed from those given other medicines. None of these studies can show that melatonin causes or prevents harm.

For children, the 2023 review, prepared for the Danish Health Authority, pooled 22 trials with 1,350 participants and found non-serious side effects about one and a half times as often as on placebo, and it found no studies at all on bone health. UK guidance for autistic children recommends considering melatonin only when sleep problems persist after a behavioural sleep plan, with a specialist involved and regular review of whether the benefits still outweigh the risks. The US agency notes that because melatonin is a hormone, effects on puberty, menstrual cycles and the hormone prolactin in children are possible but have not been confirmed. The US government's liver injury database rates melatonin an unlikely cause of liver damage. In the US melatonin is sold as a dietary supplement, while in several other countries it is available only by prescription.