Glycine for Sleep: 11 People, 2 Nights, and the Company That Sells It

Glycine has become the sleep supplement for people who think they are too evidence-literate for melatonin. It is cheap, it is an amino acid your body already makes, it tastes vaguely sweet in water, and it comes with a tidy mechanism about body temperature that sounds like real physiology rather than marketing. Three grams before bed, and the studies back it up.

The studies do exist. That part is true, and it is more than you can say for most of the shelf. What almost nobody does is open them and read the methods sections. When you do, the picture that emerges is not that glycine is a fraud. It is that an entire internet consensus has been built on top of roughly 40 people, studied by the company that manufactures the ingredient, and that the result which replicated most reliably was not about your sleep at all.

Where the glycine sleep story actually comes from

There are three original human trials. They are the source of every claim you have read, and they came from one research group.

The first, published by Inagawa and colleagues in Sleep and Biological Rhythms in 2006, gave 19 people with sleep complaints 3 g of glycine or a flavour-matched placebo an hour before bed for four consecutive nights. The outcome that improved was subjective fatigue, measured on a fatigue checklist. Not sleep architecture. Not how long they slept. How tired they said they felt.

The second is the one everybody cites. Yamadera and colleagues, in the same journal in 2007, took 11 self-reported poor sleepers aged 30 to 57, gave them the same 3 g an hour before bed, and this time hooked them up to polysomnography, the gold-standard sleep laboratory measurement. Those participants reported better sleep quality, and the machine showed shorter latency to sleep onset and shorter latency to slow-wave sleep. This is the study behind every “glycine is clinically proven to help you fall asleep faster” claim in circulation.

It is a real finding from a real crossover trial. It is also 11 people. The 2024 systematic review that catalogued this literature records the protocol as running for two consecutive nights. Neither of these papers is indexed in PubMed, which is part of why so few people who cite them have actually read them.

The trial whose methods section nobody reads

The third study is where it gets uncomfortable, and it is the one worth spending time on because it is fully open access and you can check every word.

Bannai and colleagues published it in Frontiers in Neurology in 2012. The design was clever: restrict people’s sleep to 25 percent less than their usual time in bed for three consecutive nights, give them 3 g of glycine or placebo before bed, then measure how they functioned the next day. The results are quoted everywhere. Significant reduction in fatigue on a visual analogue scale, significant improvement on the psychomotor vigilance test, and what the authors carefully described as “a tendency toward reduced sleepiness.”

Read that last phrase again. A tendency is what researchers write when the result did not reach statistical significance. The headline effect on daytime sleepiness, in the study most often used to prove glycine fights daytime sleepiness, did not clear the bar.

Now the methods. The paper states that the participants were “Ten healthy, male volunteers (30–61 years of age; average, 41.4 years) who were employed at Ajinomoto Co., Inc. as full-time weekday workers.” Ajinomoto is one of the largest amino acid manufacturers in the world. The authors’ own affiliation line reads “Frontier Research Labs, Institute for Innovation, Ajinomoto Co., Inc.”

It gets thinner. “Among the 10 participants, 1 dropped out, and another 2 who took medication for influenza and a pollen allergy during the experimental period were excluded. Statistical analysis was performed on seven subjects.” And the design is described by the authors as “a randomized single-blinded crossover trial” — single-blinded, not double-blinded.

So the study that anchors glycine’s daytime-performance claim analyzed seven men, all of them employees of the company that sells glycine, in a single-blinded design, and its sleepiness result was a non-significant trend.

None of that makes the finding false. Industry scientists produce real science, and Ajinomoto’s group did the work when nobody else was interested. But it is a very long way from what “clinically proven” is doing in the marketing copy.

The mechanism is elegant, and it was worked out in rats

The temperature explanation is the most persuasive part of the glycine pitch, and it deserves credit for being genuinely good physiology. Your core body temperature falls as you fall asleep. Glycine appears to accelerate that by dilating blood vessels near the skin so you dump heat faster.

The definitive version of this work is a 2015 paper in Neuropsychopharmacology by Kawai and colleagues, again with Ajinomoto affiliations, this time alongside Stanford’s sleep laboratory. It is a careful, convincing piece of research. Oral glycine induced non-REM sleep and shortened non-REM sleep latency while core temperature dropped. The researchers traced the effect to NMDA receptors in the shell of the suprachiasmatic nucleus, and showed that ablating that nucleus abolished the sleep-promoting and hypothermic effects entirely.

Every one of those experiments was in rats.

That matters more than it sounds. A mechanism demonstrated in a rodent brain tells you the pathway is biologically possible; it does not tell you the pathway is doing anything measurable when a human being takes a scoop of powder. And when somebody finally measured core temperature in sleeping humans given a glycine-rich supplement, they found nothing. More on that in a moment.

What happened when independent researchers looked

Three pieces of non-Ajinomoto evidence have arrived, and they pull in interesting directions.

The systematic review. In 2024, a team from the National University of Singapore and Vrije Universiteit Amsterdam published a systematic review in GeroScience covering glycine’s effects across eleven physiological systems in adult humans. Their declaration reads “No conflict of interest to declare.” On sleep, their verdict is a single sentence worth quoting exactly: “While longer-term glycine administration improved sleep in healthy populations, these studies had small sample sizes with a high risk of bias.” They also note the boundary of the whole literature — glycine had been administered to healthy populations for up to 14 days, and to people with existing illness for up to 4 months. Nobody has studied nightly glycine in healthy people beyond about two weeks.

The collagen trial, which cuts both ways. In 2024, Thomas and colleagues published a randomized crossover trial in the European Journal of Nutrition giving 13 athletic men with sleep complaints 15 g per day of glycine-rich collagen peptides an hour before bed for seven nights, with polysomnography on the final night. Their conclusion, verbatim: “CP supplementation did not influence sleep quantity, latency, or efficiency, but reduced awakenings and improved cognitive function in physically active males with sleep complaints.” Polysomnography showed fewer awakenings on the supplement, 21.3 against 29.3. But there were no differences in core temperature, which is a direct miss for the mechanism the whole glycine story rests on. Two honest caveats: collagen peptides are not free glycine, and two of the authors are employees of the collagen manufacturer.

The blend trial. A 2023 study in Medicine & Science in Sports & Exercise used an unusually rigorous double-blind deception design, so most participants did not know sleep was the outcome being measured. In 16 people it produced the strongest numbers in this entire article: sleep onset latency down by 24 minutes, total sleep time up 22 minutes, sleep efficiency up 2.4 percent. The problem for our purposes is that the supplement contained tryptophan, glycine, magnesium, tart cherry powder and L-theanine. You cannot pull glycine’s contribution out of that. It is evidence that a stack can work, not evidence about glycine.

So: one independent review saying high risk of bias, one independent-ish trial that failed to move sleep quantity or latency and failed to find the temperature effect, and one strong result that cannot be attributed to glycine. That is the whole non-manufacturer picture.

What this means if you still want to try it

Nothing above says do not try glycine. It says know what you are buying.

Use the dose from the trials, which is 3 g roughly an hour before bed. There is no human sleep data at any other dose or timing, so improvising upward buys you nothing. Glycine trials in other conditions have used far higher daily doses under medical supervision without the literature flagging acute harm, but headroom is not a reason to climb into it.

Expect the effect that replicated, not the one that sells. Across these studies, the finding that kept reappearing was about the following day: less fatigue, better psychomotor vigilance. The sleep architecture result came from a single 11-person laboratory study. If you take glycine for a week and cannot tell any difference in how quickly you fall asleep, you are not doing it wrong; you are getting the result the evidence should have led you to expect.

Do not confuse it with magnesium glycinate. This trips up a lot of people. In magnesium bisglycinate, elemental magnesium is only about 14 percent of the compound’s molecular weight and the two glycine molecules make up most of the remainder. Run the arithmetic on a typical 200 mg elemental magnesium dose and you land somewhere near 1.2 g of glycine — meaningfully short of the 3 g every trial used. If you are choosing between magnesium glycinate and citrate, choose on the magnesium, not on a glycine dose you are not actually getting. The same goes for stacking it with magnesium and melatonin: three weak effects do not add up to one strong one.

Collagen at bedtime is a different question. Collagen is genuinely glycine-rich, which is why the trial above exists, but 15 g of collagen peptides is not 3 g of free glycine and the results were not the same. If that is your interest, the case for taking collagen at night stands or falls on its own evidence.

Run it as an experiment, not a purchase. Give it two weeks, which is also the outer limit of what has been studied in healthy people. Change one thing at a time. Track how you feel the next afternoon, not just how fast you dropped off, because next-afternoon is where the signal has actually been.

When a supplement is the wrong tool entirely

This is the part that matters more than the dose.

Persistent insomnia is a medical problem with a first-line treatment, and the treatment is not a powder. Cognitive behavioural therapy for insomnia is what sleep medicine recommends before medication, and it outperforms supplements by a margin that makes this whole discussion look small. The American Academy of Sleep Medicine’s guideline on pharmacologic treatment of chronic insomnia evaluates individual agents including suvorexant, eszopiclone, zolpidem, zaleplon, triazolam, temazepam, ramelteon and doxepin. Glycine does not appear among them.

That same guideline contains a line that reads, unintentionally, like a summary of this entire article. Explaining why it downgraded the quality of evidence even for approved drugs, the task force wrote that doing so “is predictable in GRADE, due to the funding source for most pharmacological clinical trials and the attendant risk of publication bias; the relatively small number of eligible trials for each individual agent.” Small trials, funded by interested parties. Sleep medicine already knows to discount for exactly this. Supplement marketing does not.

And some causes of bad sleep are not insomnia. Loud snoring with pauses in breathing, waking unrefreshed no matter how long you were in bed, morning headaches, or an irresistible urge to move your legs at night all point somewhere else, and no amino acid addresses any of them. If your sleep has been broken for more than a few weeks, that is a conversation with a doctor, not a shopping decision. It is also worth checking whether the answer is simpler than a supplement — the evidence ranking for sleep supplements generally is humbling, melatonin has a narrower job than most people think, and L-theanine is a calmer comparison case with its own limits.

The honest bottom line

Glycine’s sleep reputation rests on three small trials from one research group at an amino acid manufacturer, totalling 40 enrolled participants. The most-cited of them studied 11 people over two nights. The most-quoted daytime result came from seven analyzed men who worked for the company, in a single-blinded design, with the sleepiness finding falling short of significance. The mechanism is well-demonstrated in rats and unconfirmed in humans, and the one trial that measured core temperature during human sleep found no difference.

The independent review that assessed all of it landed on “small sample sizes with a high risk of bias,” and that is the fairest sentence anyone has written about glycine and sleep.

None of this makes glycine a bad thing to try. It is inexpensive, it appears safe at 3 g, and there is a real signal in there about feeling less wrecked the next day. What it is not is settled. If you buy it, buy it as a two-week experiment with a plausible mechanism and thin evidence — and if your sleep is genuinely broken, spend the effort on the thing that actually has the trial data behind it.

This article is for general information and is not medical advice. Talk to a qualified healthcare professional before starting any supplement, especially if you take medication, are pregnant or breastfeeding, or have a medical condition.