In a 2025 trial in 155 adults who reported poor sleep, the group taking 250 mg of magnesium a day as bisglycinate scored 3.9 points lower on the Insomnia Severity Index after 4 weeks, against 2.3 points lower on placebo; the authors rated the difference small. That dose arrives with about 1.5 g of glycine, half of the 3 g used in a 2012 glycine study.
What the sleep studies measured
Of the 5 magnesium sleep trials below, published 2002 to 2026, only 1 used bisglycinate, and it measured sleep by questionnaire alone. Its abstract does not describe what the placebo contained, so the trial cannot separate the magnesium from the roughly 1.5 g of glycine that came with it.
A 2012 trial in 46 older adults with insomnia measured insomnia severity, sleep logs and blood hormones after 500 mg of magnesium a day (the abstract does not name the salt) for 8 weeks: insomnia scores, time to fall asleep and sleep efficiency moved more than on placebo, while total sleep time did not differ (P = 0.37).
A 2021 meta-analysis of 3 trials in 151 older adults measured time to fall asleep: it was 17.36 minutes shorter on magnesium than on placebo. Total sleep time was 16.06 minutes longer, which was not significant, and the authors graded the evidence low to very low quality.
| Year · journal · PMID | People | Salt and dose | Length | Measured | Result | Limitation |
|---|---|---|---|---|---|---|
| 2025 · Nat Sci Sleep · 40918053 | 155 adults aged 18–65 with poor sleep | Bisglycinate, 250 mg magnesium a day | Primary result at 4 weeks | Insomnia Severity Index | −3.9 vs −2.3 on placebo (p = 0.049); other psychological scores did not differ | Questionnaires only; small effect (d = 0.2) |
| 2012 · J Res Med Sci · 23853635 | 46 older adults with insomnia | 500 mg a day; salt not named | 8 weeks | Insomnia score, sleep log, melatonin, cortisol | Insomnia score, sleep onset and efficiency favored magnesium; total sleep time did not differ | Small; dose above the 350 mg supplement limit |
| 2021 · BMC Complement Med Ther · 33865376 | 3 trials, 151 older adults | Oral magnesium, doses varied | Varied | Sleep onset, total sleep time | Onset 17.36 min shorter; total sleep +16.06 min, not significant | Low to very low quality evidence |
| 2002 · Pharmacopsychiatry · 12163983 | 12 adults aged 60–80 | Effervescent tablets rising to 30 mmol a day (about 729 mg); salt not named | 20 days per arm, crossover | Sleep EEG, night hormones | Slow-wave sleep 16.5 vs 10.1 min; cortisol lower | 12 people; about twice the supplement limit |
| 2011 · J Am Geriatr Soc · 21226679 | 43 care-home residents, mean age 78 | 225 mg magnesium with 5 mg melatonin and 11.25 mg zinc | 8 weeks | Pittsburgh Sleep Quality Index | Change differed from placebo by 6.8 points | Magnesium cannot be separated from melatonin |
| 2026 · Front Nutr · 41601871 | 100 adults aged 18–45 with dissatisfied sleep | L-threonate, 2 g a day (about 165 mg magnesium by formula) | 6 weeks | Sleep questionnaires, Oura ring | Sleep-related impairment favored threonate (p = 0.043); sleep disturbance and ring sleep data did not differ | Funded by the ingredient's owner |
| 2012 · Front Neurol · 22529837 | Healthy adults, sleep cut by 25% for 3 nights; number not in abstract | Glycine 3 g, no magnesium | 3 nights | Daytime sleepiness, fatigue, vigilance test | Fatigue lower; sleepiness only a trend | Authors work for a glycine maker |
| 2023 · Biol Trace Elem Res · 35184264 | 9 studies, 7,582 people | Diet or supplements | Varied | Sleep quality and duration | Observational studies: linked; trials: contradictory | Mixes surveys with trials |
| 2024 · Cureus · 38817505 | 15 trials, 8 on sleep | Varied | Varied | Sleep and anxiety scores | 5 of 8 sleep studies positive, 2 null, 1 mixed | Small samples, mixed forms and doses |
The arithmetic of a glycinate dose
Magnesium bisglycinate, Mg(C₂H₄NO₂)₂, is 11.8–14.1% magnesium by weight depending on bound water, so 200 mg of magnesium needs 1,418–1,691 mg of the salt and carries about 1.2 g of glycine.
| Magnesium a day | Bisglycinate needed | Size 00 capsules (730 mg each) | Glycine that comes with it |
|---|---|---|---|
| 200 mg | 1,418–1,691 mg | 2–3 | 1.2 g |
| 250 mg (2025 trial) | 1,773–2,113 mg | 3 | 1.5 g |
| 350 mg (supplement upper limit) | 2,482–2,959 mg | 4–5 | 2.2 g |
| 400 mg (common label figure) | 2,837–3,381 mg | 4–5 | 2.5 g |
Capsule counts assume pure salt at 0.8 g/mL with no filler. Getting the 3 g of glycine from the 2012 study out of bisglycinate would take about 486 mg of magnesium, above the 350 mg limit. Bound water does not change the glycine per milligram of magnesium, but a label that mixes in oxide delivers less glycine for the same magnesium.
On the shelf, 73 of 126 on-market NIH DSLD labels that name glycinate name it as the only magnesium salt, and 14 also name magnesium oxide (snapshot 2026-08-10). 3 of 25 capsule labels naming glycinate alone declare more magnesium per capsule than pure glycinate could fit even in a size 000 capsule; the label audit lists them.
Best magnesium glycinate for sleep, by three measurements
Within glycinate, the cheapest ranked label is Amazon Elements Chelated Magnesium 270 mg at $0.0666 per 100 mg of magnesium, against a median of $0.1889 across 17 ranked offers (prices observed 2026-10-26 to 2026-11-02). "Best" here names a measurement, not a sleep result: no trial compared brands.
| Measurement | Label | Number |
|---|---|---|
| Lowest price per 100 mg, within glycinate | Amazon Elements Chelated Magnesium 270 mg | $0.0666; $4.00 for 30 days at 200 mg |
| Lowest price with no oxide or other salt named | Amazon Elements Chelated Magnesium 270 mg | $0.0666 per 100 mg |
| Fewest capsules or tablets for 200 mg | Amazon Elements Chelated Magnesium 270 mg | 2 a day; 3 for 400 mg |
| Same-salt median | All ranked glycinate offers | $0.1889; $11.33 for 30 days at 200 mg |
Prices are compared only within glycinate. Threonate, the salt in the 2026 trial, has its own page.
Safety
The upper limit for magnesium from supplements is 350 mg a day for adults, set in 1997 and listed by the NIH Office of Dietary Supplements; magnesium from food does not count toward it. Two trials above gave more under study supervision: 500 mg and about 729 mg a day. In the cramp trials pooled by Cochrane in 2020, gut effects such as diarrhoea were reported by 11–37% of people on oral magnesium, against 10–14% on control (32956536). Talk to a clinician if you have kidney disease, take prescription medicines or are pregnant, or if poor sleep has lasted for weeks.