People take melatonin because they heard it helps you fall asleep. They take valerian root because the bottle says “natural sleep support.” They take magnesium because someone on a wellness podcast made it sound obvious. And here’s the thing — these supplements actually do something. They just don’t do very much.
About 7 minutes faster to sleep with melatonin. About 8 more minutes of total sleep. Maybe 17 minutes with magnesium, though the evidence for that one is built on three small studies — and the authors of the best available review called it “substandard for physicians to make well-informed recommendations.”
That’s the honest version. It’s not nothing. But it’s the kind of nothing-plus-a-little that supplement marketing rounds up into a sleep revolution.
Melatonin: the numbers are real but small#
Of the three, melatonin is the most studied. A 2013 meta-analysis in PLoS ONE pulled together 19 studies with more than 1,600 people (Ferracioli-Oda et al.). The results: melatonin reduced the time it takes to fall asleep by about 7 minutes on average, and added about 8 minutes of total sleep. Sleep quality improved, but the effect was modest — the standardized mean difference was 0.22.
For context, prescription sleep medications do more. Benzodiazepines cut sleep latency by 10 to 20 minutes. Z-drugs like Ambien reduce it by 13 to 17 minutes. Melatonin’s 7 minutes is real and measurable. It’s also roughly a third to half of what a prescription drug does.
And there’s a catch. That total-sleep-time gain of 8 minutes showed up when people reported how long they slept. When researchers measured sleep objectively with polysomnography or actigraphy, the improvement disappeared. People felt like they’d slept more. Whether they actually did is less clear.
Melatonin isn’t a sedative. It’s a hormone your pineal gland produces in darkness, and its job is signaling your circadian clock — not knocking you out. The strongest evidence is for circadian rhythm disorders (jet lag, delayed sleep-wake phase disorder), not general insomnia. The NIH’s National Center for Complementary and Integrative Health says it “may help” with those specific uses but is uncertain whether benefits outweigh possible harms for other sleep problems.
Melatonin also doesn’t play nicely with other medications. If you take blood thinners or seizure meds, birth control pills or blood pressure drugs, diabetes medications or immunosuppressants — Mayo Clinic flags all of them. And one piece of practical advice that matters if you’re considering it: do not drive or use machinery within 5 hours of taking melatonin.
Higher doses over longer periods do seem to produce somewhat larger effects, and a 2024 dose-response review suggested 4 mg about 3 hours before bedtime might work better than the common 2-mg-at-bedtime routine. Which would be useful — if you could trust what’s in the bottle. That’s the problem.
You don’t know what you’re actually taking#
Even if melatonin works a little, it only works when the dose you take is the dose you meant to take. The evidence says it probably isn’t.
A 2017 analysis of 31 melatonin supplements in the Journal of Clinical Sleep Medicine (Erland & Saxena) found that more than 71% didn’t match their label claims within a 10% margin. One product had nearly five times the advertised dose. Another had barely any melatonin at all. Lot-to-lot variability hit 465% — same brand, different batches, wildly different content.
Then came the serotonin finding. About 26% of products — 8 of 30 — contained unlabeled serotonin at 1 to 75 micrograms. If you’re taking an SSRI antidepressant or tramadol, adding unlabeled serotonin could, in theory, trigger serotonin syndrome, a potentially dangerous condition.
A 2023 JAMA study made things worse. Researchers tested 25 melatonin gummy products. Twenty-two — that’s 88% — were inaccurately labeled. Actual content ranged from 74% to 347% of the label. Only three products fell within 10% of what they claimed. The Washington Post, CBS News, and NPR all covered it. The most popular form of the most popular sleep supplement is basically a lottery.
The FDA keeps a public list of sleep supplements found to contain hidden drug ingredients — products like U-Dream Full Night and U-Dream Lite that contained undeclared pharmaceuticals. And the agency has urged retailers including Amazon to stop selling them.
Here’s the thing: melatonin isn’t FDA-approved to treat any condition. It’s sold as a dietary supplement, which means the FDA doesn’t review it for safety, effectiveness, or label accuracy before it hits shelves. Nobody checks the bottle before you buy it. That’s why the quality-control failures documented above aren’t a scandal — they’re just how the system works. When there’s no pre-market review, the only enforcement happens after people get hurt.
And the consequences aren’t abstract. Between 2012 and 2021, pediatric melatonin ingestions reported to U.S. poison control centers increased 530%, according to the CDC — from about 8,300 to more than 52,500 per year. Melatonin became the most frequently ingested substance by children in poison control data. The CDC noted that the most variable dosing was found in chewable formulations — the kind most likely to be ingested by children. Two children under age 2 died, though the CDC couldn’t confirm melatonin as the cause in those cases. When a hormone is sold in candy form with doses ranging from a quarter to triple the label, bad things happen.
Valerian: mixed evidence, opposing experts#
Valerian root has been used for sleep since ancient Greece — Hippocrates and Galen both wrote about it. That’s real traditional-use history. The clinical evidence, though, is a mess.
A 2006 meta-analysis in the American Journal of Medicine pooled 16 studies and found valerian users were about 80% more likely to report improved sleep (Bent et al.). Sounds solid — until you read the authors’ caveat about publication bias, meaning negative valerian studies may not have been published at all, inflating the positive signal.
A 2020 review covering 60 studies and nearly 7,000 people (Shinjyo et al.) sharpened one useful point: whole valerian root at 450 to 1,410 mg per day for four to eight weeks produced more consistent sleep improvements than standardized extracts. The catch is consumers can’t tell whether they’re buying whole root or an extract, and even if they could, there’s no guarantee the label is accurate.
The NIH Office of Dietary Supplements calls the evidence “inconclusive.” The Mayo Clinic says results are inconsistent and notes valerian often needs two weeks before any effect appears. The American Academy of Sleep Medicine recommends against valerian for both sleep onset and sleep maintenance insomnia. But the European Medicines Agency recognizes valerian for mild nervous tension and sleep disorders. Different expert bodies, same evidence, opposite conclusions. Nobody’s hiding the contradiction.
Magnesium: the thinnest evidence of the three#
Magnesium feels more trustworthy than the others — it doesn’t sound like a drug or herb, and magnesium deficiency is a real thing. The mechanism makes intuitive sense: magnesium acts on NMDA and GABA receptors in ways that could reduce excitatory signaling and promote relaxation.
But the clinical evidence is strikingly thin. The best review available, published in BMC Complementary Medicine and Therapies in 2021 (Mah & Pitre), could only find three randomized controlled trials — three — totaling 151 older adults. That’s the entire evidence base. The latency reduction looked meaningful at about 17 minutes and was statistically significant, but the total sleep time gain of about 16 minutes was not significant. All three trials carried moderate-to-high risk of bias.
But a single 2025 trial added one more data point showing modest improvement in insomnia severity scores. It doesn’t change the picture. The 2021 authors didn’t sugarcoat it: the evidence is “substandard for physicians to make well-informed recommendations.”
What the experts say (and why they disagree)#
Here’s where it gets genuinely interesting.
The American Academy of Sleep Medicine’s 2017 guideline recommends against melatonin and valerian for chronic insomnia in adults. Every recommendation was rated weak — the evidence wasn’t strong enough for a confident yes or no — but the direction was clear. First-line treatment: cognitive behavioral therapy for insomnia (CBT-I). The VA/Department of Defense joint guideline says the same.
But the American Academy of Family Physicians, representing more than 130,000 family doctors, recognizes melatonin as first-line pharmacological therapy for insomnia. Not a fringe view — the official position.
Sleep medicine specialists say no. Family doctors say yes, as a first step. Same evidence, different interpretation of what “weak but real” means in practice.
CBT-I, meanwhile, is the one thing every major guideline agrees on. It’s a structured program — usually 4 to 8 sessions — targeting the thoughts and behaviors that maintain insomnia. It’s not sleep hygiene tips. It’s not chamomile tea. It’s evidence-backed, and it doesn’t come in a bottle with a label that might be wrong.
How to think about it#
None of this means sleep supplements are useless. It means they’re much smaller tools than the wellness industry implies.
Melatonin probably shaves a few minutes off sleep onset. That’s genuinely useful for jet lag or delayed sleep phase. For chronic insomnia? Less so. Valerian might help some people, but the evidence is inconsistent and long-term safety data is thin. If you have liver disease, the Mayo Clinic says avoid it. If you take sedatives, alcohol, or other CNS depressants, the additive effects are real. Magnesium’s early numbers are interesting but the evidence isn’t there yet, and the most common side effect at higher doses is diarrhea — which doesn’t help anyone sleep.
And there are groups that should be especially careful. If you’re pregnant or breastfeeding, melatonin and valerian both come with insufficient safety data — melatonin’s hormonal effects on development are unknown, and valerian hasn’t been studied in pregnancy at all. If you have impaired kidney function, magnesium supplements are a no-go without medical supervision — your kidneys regulate magnesium levels, and toxicity risk jumps when they’re not working right.
And the supplement quality problem hangs over all of this. When 71% of products are mislabeled, 88% of gummies are inaccurate, and one in four contains an unlabeled neurotransmitter, the numbers on the bottle are fiction. You’re guessing.
If you’re going to use these supplements anyway, look for third-party certification (USP, NSF, or ConsumerLab), stick with simpler formulations, and store everything where children can’t reach it — especially gummies, which look and taste like candy because that’s what they are.
And if sleep is a persistent problem — a pattern, not an occasional bad night — the most evidence-supported thing you can do isn’t in the supplement aisle. It’s talking to a doctor, ruling out conditions like sleep apnea or restless leg syndrome, and asking whether CBT-I is available. The supplements are the appetizer. The behavioral stuff is the main course.
