In short: for most people, a well-absorbed organic form is the best choice, above all citrate or glycinate. Both are usually absorbed better in studies than magnesium oxide. Citrate is the best researched and affordable; glycinate is considered especially gentle on the stomach. You get even more out of it if you split your daily amount across two meals and take it regularly.
Key points
- Citrate and glycinate are a good choice: a systematic review puts organic forms ahead for absorption.
- Citrate is the best researched: in several comparisons, more magnesium reached the body from citrate than from oxide.
- Small portions go further: your body absorbs a much bigger share from small portions than from large ones.
- Glycinate for the evening: in a study of 155 adults, sleep problems decreased more with bisglycinate than with a placebo.
- The magnesium content decides the capsule count: oxide is 60% magnesium, threonate only just over 8%.
Contents
- Which magnesium is best?
- All magnesium forms in one table
- Organic or inorganic magnesium?
- Which magnesium is absorbed best?
- Magnesium citrate or magnesium oxide?
- The forms in detail
- Which magnesium for what?
- Magnesium complex: is a blend better?
- How to spot a good supplement
- Frequently asked questions
- Limits of the research and all sources
Which magnesium is best?
For most people, it comes down to a well-researched organic form such as citrate or glycinate. If you have a sensitive stomach, try glycinate. If price matters to you, citrate is a good fit. Speciality forms such as threonate are worth a look if you're interested in a specific question that has been studied.
Just as important as the form: the best magnesium is the one you tolerate well, take in a sensible amount and use regularly.
What the studies show: a systematic review of 14 comparative studies found that organic forms are absorbed better than inorganic ones such as oxide. All the forms tested kept magnesium levels normal in healthy people.1 A review from Hannover highlights how much the amount per portion and your existing magnesium status matter.2
All magnesium forms in one table
The table shows the common forms at a glance. We calculated the "magnesium content" column ourselves from the molar masses (the maths is below). The "strength of evidence" column tells you how much weight to give the claims.
| Form | Magnesium content (our calculation) | Absorption in human studies | Tolerability | Typical use | Strength of evidence |
|---|---|---|---|---|---|
| Magnesium oxide (inorganic) | 60.3% | usually worse than citrate, but not in every study5–12 | laxative in large amounts; in one long-term study, fewer digestive complaints than citrate9, 17 | cheap tablets with a lot of magnesium | well studied, mixed results |
| Magnesium citrate (organic) | 16.2% (anhydrous) | better than oxide in several comparisons5–7, 9 | usually good, can loosen stools | standard form in powders and capsules | good |
| Magnesium glycinate / bisglycinate (organic) | 14.1% | better than oxide according to a review1, 13, 14 | considered gentle on the stomach; one small study: better tolerated than oxide14 | sensitive stomach, evening use | medium |
| Magnesium malate (organic) | 15.5% | good results in rats; human comparisons still missing18 | no comparative data | sport, "energy" products | animal studies so far |
| Magnesium taurate (organic) | 8.9% | so far from animal studies18, 31 | no comparative data | marketed for heart and circulation | animal studies so far |
| Magnesium L-threonate (organic) | 8.3% | more magnesium in the brain (in rats)21 | well tolerated in studies23, 24 | memory, sleep | early human studies |
| Magnesium chloride (inorganic) | 25.5% (anhydrous), 12.0% (hexahydrate) | about as good as organic forms, better than oxide8, 12, 28 | tastes bitter | solutions, "magnesium oil" | medium |
| Magnesium carbonate (inorganic) | 28.8% | no rise in a small study with effervescent tablets11 | no comparative data | effervescent tablets | little data |
| Magnesium sulfate (inorganic) | 20.2% (anhydrous), 9.9% (heptahydrate) | blood level similar to citrate, less in urine9 | more digestive complaints than placebo in one study9 | Epsom salt, bath salt | medium |
| Sea magnesium (usually hydroxide or oxide from seawater) | depends on product; hydroxide 41.7% | one small study: rise compared with placebo26 | no comparative data | "natural" magnesium | one small study |
How we calculate this: the magnesium content comes from the weight of the atoms. Magnesium weighs 24.3 units. The whole compound weighs 40.3 units for oxide, 172.4 for bisglycinate and 294.5 for threonate (molar masses from PubChem).4 So: 24.3 ÷ 40.3 = 60.3% for oxide, and 24.3 ÷ 294.5 = 8.3% for threonate. For citrate, three magnesium sit in one unit of 451.1: 3 × 24.3 ÷ 451.1 = 16.2%.
If the powder contains bound water (a hydrate), the share is lower. That's why we give both figures for chloride and sulfate.
Organic or inorganic magnesium?
Organic forms are, on average, better absorbed. That's why citrate and glycinate are a good choice for most people. "Organic" doesn't mean "organically farmed" here. It simply describes what the magnesium is bound to: in organic forms, a carbon-containing acid or amino acid, such as citric acid, glycine or malic acid. Inorganic forms are oxide, hydroxide, carbonate, chloride and sulfate.
Two things are worth knowing. Some inorganic forms also do well, above all chloride. And organic forms contain less magnesium per gram, so you need a bit more powder or more capsules for the same amount.
What the studies show: the systematic review puts organic forms ahead.1 In a US study, chloride, lactate and aspartate were absorbed equally well, all clearly better than oxide. The authors conclude that, depending on the product, inorganic salts can be just as good as organic ones.8 A 2025 review rates the bioavailability of citrate and chloride as good and that of oxide as poor.28
Which magnesium is absorbed best?
The best-documented edge is citrate's advantage over oxide. Chloride and glycinate also do well in studies. But you make the biggest difference yourself: small portions are absorbed far better than one large one.
"Bioavailability" means the share that makes it from the gut into the blood. It's usually measured via urine: the body excretes much of whatever extra it absorbs.
What the studies show: in a study using magnesium with a meal, the absorbed share fell from 65% at the smallest amount to 11% at the largest.3 Supplements also vary widely in how well they dissolve. Of 15 commercially available products, the one that performed worst in the lab also performed markedly worse in people.15 In animal studies with ten forms, absorption ranged between 50 and 67%, with organic forms slightly ahead.16
Our take: with a well-soluble form such as citrate, glycinate or chloride, you're on the safe side for absorption. You don't need to agonise over which of these three to pick: take the one you tolerate best. The biggest gain comes from splitting your daily amount across two meals. Which of the well-soluble forms comes out ahead in the long run is something larger studies still need to show.
Magnesium citrate or magnesium oxide?
If you're only choosing between these two: citrate is absorbed better in most studies and is therefore the first choice. Oxide scores with a lot of magnesium per tablet and a low price.
Why citrate? Oxide barely dissolves in water and only partly in stomach acid. Citrate already dissolves well in water. If you produce little stomach acid, for instance because of acid-reducing medicines, you're therefore better off with citrate.
What the studies show: in a lab test, oxide dissolved only 43% even with plenty of stomach acid, while citrate already dissolved 55% in plain water. Considerably more magnesium showed up in participants' urine after citrate.5 In a 60-day study of 46 healthy people, citrate and an amino acid chelate did better than oxide; oxide didn't differ from placebo.6 A study of 20 men reached the same conclusion.7 In a US study, only around 4% was absorbed from oxide.8
Our take: if you're choosing for the first time, go for citrate. Most studies support it, especially the careful ones with urine measurements. If you already take oxide and tolerate it well, you don't need to switch. Over a period of weeks, the body balances out a lot: in the longest comparative study, blood levels rose equally with both forms.9 You'll find more on citrate in our article on magnesium dicitrate.
The forms in detail
Magnesium oxide
Magnesium oxide is magnesium combined with oxygen. It's the cheapest form and has the highest content at 60.3%. A small tablet is therefore enough for a full daily portion. It dissolves poorly, though, and a large part stays in the gut, where it draws in water. That's exactly why magnesium oxide is also used as a laxative in larger amounts, especially in East Asia.17
What the studies show: most comparisons put oxide behind citrate; some find no difference or the reverse.5–12 In the longest comparative study, the oxide group actually reported fewer digestive complaints than the citrate group.9
Magnesium citrate
Magnesium citrate is magnesium bound to citric acid. Usually trimagnesium dicitrate is meant, with 16.2% magnesium in the anhydrous form. It dissolves well and is the most studied organic form. In larger amounts, it can loosen the stool.
What the studies show: citrate was absorbed better than oxide in several studies.5–7, 9 You'll find the details in our article on magnesium dicitrate.
Magnesium glycinate (bisglycinate)
Here, magnesium is bound to two molecules of the amino acid glycine. The form is considered especially gentle on the stomach and is popular for evening use. In the shops, "glycinate" and "bisglycinate" refer to the same compound. Pure bisglycinate contains no more than 14.1% magnesium.
What the studies show: in patients after bowel surgery, glycinate was tolerated better than oxide. Among the patients with the poorest absorption, glycinate also came out ahead for absorption.14 In a study of 155 adults, sleep problems decreased more with 250 mg of magnesium as bisglycinate than with a placebo.30 We've brought together all the studies, the authenticity check and the sleep study in our article Magnesium glycinate and bisglycinate.
Magnesium malate
Magnesium malate is magnesium bound to malic acid. It contains 15.5% magnesium. It's popular with athletes and in "energy" products because malic acid occurs in the energy metabolism of cells.
What the studies show: in rats, blood magnesium levels stayed elevated longest after malate.18 Comparative studies in humans are still missing.
Magnesium taurate
Magnesium taurate combines magnesium with taurine, an aminosulfonic acid. It contains 8.9% magnesium. For 250 mg of magnesium, you therefore need around 2.8 g of powder (our calculation).
What the studies show: the claims about heart and circulation come from animal studies.31 In a rat study, magnesium acetyltaurate, a related compound, accumulated well in the brain.18 These are early indications that human studies still need to confirm.
Magnesium L-threonate
Threonate is a breakdown product of vitamin C. The compound was developed to get more magnesium into the brain. It contains 8.3% magnesium: 2 g of threonate provides at most around 165 mg of magnesium (our calculation).
What the studies show: in rats, brain magnesium rose with threonate, and the animals learned better.21 In humans, there are three controlled studies. In one with 100 adults, the threonate group did better on memory tests than the placebo group after six weeks; sleep improved in some respects.24 In a study of 80 adults with sleep problems, the threonate group reported better sleep and more daytime energy.23 A third study tested a blend with other active ingredients.22 All three were paid for by the manufacturer or distributor of the raw material. You'll find more in our article on magnesium L-threonate.
Magnesium chloride
Magnesium chloride is an inorganic salt that dissolves very well and is well absorbed. It tastes bitter and is therefore often sold as a solution or in tablets. Sold as "magnesium oil", a concentrated chloride solution is sprayed onto the skin.
What the studies show: for absorption through the skin, a review found no solid evidence.25 Swallowed, however, chloride is a good choice: in a US study, it was absorbed just as well as lactate and aspartate, clearly better than oxide.8
Magnesium carbonate
Magnesium carbonate contains 28.8% magnesium and is often found in effervescent tablets. There's still little data on its absorption in humans.
What the studies show: in 61 young women who took carbonate effervescent tablets for ten days, blood magnesium didn't rise.11
Magnesium sulfate (Epsom salt)
You may know magnesium sulfate as Epsom salt. As the hydrated salt, it contains 9.9% magnesium. In larger amounts, it acts as a laxative, which is why it plays almost no role as a dietary supplement.
What the studies show: in the 24-week study, blood levels rose with sulfate just as much as with citrate. The sulfate group, however, reported digestive complaints more often than the placebo group.9
Sea magnesium
"Sea magnesium" isn't a compound of its own. Magnesium from seawater is usually obtained as hydroxide or oxide, sometimes processed further. The label tells you which form ends up inside.
What the studies show: in a small study of 20 people, two seawater-derived supplements increased urinary magnesium excretion compared with placebo, with no difference between them.26 A cell study found one seawater-derived product to be about as good as chloride.27
Which magnesium for what?
To cover your basic needs, any well-absorbed form will do. For some situations, though, there are sensible preferences. The table shows what we derive from the evidence.
| If you … | obvious choice | why |
|---|---|---|
| just want to top up your intake | citrate | best researched, well absorbed, affordable |
| have a sensitive stomach | glycinate, with food | considered gentle on the stomach; better tolerated than oxide in one study14 |
| want to swallow fewer tablets | oxide or citrate | oxide has the highest content; citrate is the compromise |
| want to take it in the evening | glycinate | one study of 155 adults found fewer sleep problems than with a placebo30 |
| are specifically interested in memory | threonate | better memory tests in human studies22, 24 |
| play sport | any well-tolerated form | Magnesium contributes to normal muscle function. |
| care about your bones | any well-absorbed form | Magnesium contributes to the maintenance of normal bones. |
All forms deliver the same magnesium. The officially approved effects therefore apply to every form:
- Magnesium contributes to a reduction of tiredness and fatigue.
- Magnesium contributes to normal muscle function.
- Magnesium contributes to normal functioning of the nervous system.
- Magnesium contributes to normal psychological function.
- Magnesium contributes to the maintenance of normal bones.
Magnesium complex: is a blend better?
A magnesium complex is a blend of several forms, for example citrate, glycinate and malate. A single well-absorbed form is just as good a choice. With a blend, what matters most is that you know what's inside.
If the pack states how much magnesium comes from which form, you can check this. Many blends only state the total amount of magnesium, not the share from each form. In that case, you don't know whether most of it comes from cheap oxide.
How to spot a good supplement
The form on the front is only the start. Three details on the back reveal more.
How to do it:
- Amount of magnesium per daily serving: this means the pure magnesium, not the compound. 1,000 mg of bisglycinate contains only around 140 mg of magnesium.
- Check the ratio: divide the amount of magnesium by the amount of the compound and compare it with the table above. If the figure is clearly above what's chemically possible, another form has been mixed in. For bisglycinate, our authenticity check does the maths for you.
- Price per magnesium: compare prices per 100 mg of magnesium, not per capsule. Forms with a low content, such as threonate, need much more powder for the same amount.
For the amount, the EU recommendation helps: from supplements, adults shouldn't take more than 250 mg of magnesium a day unless a doctor advises otherwise.29 Larger amounts are best split across two meals. How much you need in total is covered in How much magnesium per day?, and when best to take it in When should you take magnesium?
Important: if your kidney function is reduced, magnesium can build up in the body, regardless of the form.32 In that case, talk to your GP first. With thyroid hormone, certain antibiotics and osteoporosis medicines, you need to leave a gap of a few hours. Details are in When should you take magnesium?
Frequently asked questions
Which magnesium is absorbed best? The best-documented finding is that citrate is absorbed better than oxide.5–7 Glycinate and chloride also do well in studies.1, 8 Among these well-soluble forms, you can choose based on tolerability. Small portions improve absorption further.3
Magnesium malate or magnesium citrate: which is better? Citrate is far better studied in humans and is therefore the obvious choice. For malate, there are good absorption figures from animal studies.18
Which magnesium is best for bones? Any well-absorbed form, such as citrate or glycinate. Magnesium contributes to the maintenance of normal bones. That applies to every form, because they all deliver the same magnesium. If you take osteoporosis medicines, leave a gap of a few hours before or after your magnesium.
Does magnesium through the skin do anything? Swallowing is the reliable way to take magnesium. The skin is a good barrier, and a review found no convincing evidence for meaningful absorption via sprays, oils or baths.25
Is magnesium from food worse than from supplements? No. In one study, magnesium from almonds was absorbed just as well as from a well-soluble magnesium salt.3 Good sources are nuts, seeds, wholegrains and legumes.
Which magnesium for high blood pressure? Researchers have studied magnesium and blood pressure in many trials. A 2025 meta-analysis evaluated 38 controlled studies with 2,709 participants. On average, blood pressure fell by 2.8/2.1 mmHg more with magnesium than with placebo. The difference was clearer in people taking blood pressure medication and in people with low magnesium levels.33 We don't know of a form that stands out here; the claims about taurate come from animal studies.31 High blood pressure should be managed by a doctor. If you take blood pressure medication, talk to your GP before starting magnesium.
Which magnesium for histamine intolerance? Look for a short ingredients list without unnecessary additives, and when in doubt, talk to your GP. We haven't found studies on magnesium forms specifically for histamine intolerance.
Limits of the research and all sources
Most comparative studies are small, last only a day and measure blood levels, which the body regulates tightly. Many results apply to one specific product, not to every form of the same name. Longer studies find smaller differences: in the longest comparative study to date, 164 adults took 450 mg of magnesium a day for 24 weeks. Blood levels rose equally with citrate, oxide and sulfate; only in urine did citrate come out ahead. None of the three forms changed blood pressure.9
Other studies don't always see citrate ahead either. In a study from Israel, cellular magnesium rose more with oxide than with citrate, though at almost double the amount.10 In 61 young women, oxide performed best after ten days.11 In another small study, there was no difference between citrate, hydroxide and chloride.12 For glycinate, one study found no measurable rise in blood levels after a single dose of bisglycinate; it was paid for by a manufacturer of a competing form.13
Human studies on malate and taurate are almost entirely missing. A combination of magnesium and malic acid showed no effect against placebo in the blinded phase of a small study in people with fibromyalgia.19 An evaluation of all the reviews concluded that the combination makes little or no difference for pain.20 For threonate, there are only studies paid for by the raw material manufacturer, and the studies on sea magnesium come from researchers linked to the manufacturer.26, 27 We haven't found studies comparing blends with a single good form. The blood pressure meta-analysis found large differences between the individual studies, and the effect wasn't clear-cut at normal blood pressure.33 Several comparisons were funded by manufacturers of specific forms; this is noted in the table for each study.
Show all 33 sources
| # | Study | Who was studied | Result | Strength of evidence |
|---|---|---|---|---|
| 1 | Pardo et al. 2021, Nutrition | 14 comparative studies | organic forms better than inorganic, share falls with amount; all forms kept levels normal in healthy people | systematic review |
| 2 | Schuchardt & Hahn 2017, Curr Nutr Food Sci | review on absorption | salt form less important than thought; amount and existing status matter more | review article |
| 3 | Fine et al. 1991, J Clin Invest | healthy people, increasing amounts with a meal | absorbed share falls from 65% to 11%; almonds as good as dissolved salt | controlled measurement study |
| 4 | PubChem: Oxide 14792, Citrate 6099959, Bisglycinate 84645, Malate 164748, Taurate 13343447, Threonate 45489777, Chloride 5360315, Chloride hexahydrate 24644, Carbonate 11029, Sulfate 24083, Sulfate heptahydrate 24843, Hydroxide 73981 | chemical database | molar masses underlying our content calculation | reference data |
| 5 | Lindberg et al. 1990, J Am Coll Nutr | lab solubility + healthy people | citrate more soluble, more magnesium in urine than after oxide | lab + small study |
| 6 | Walker et al. 2003, Magnes Res | 46 healthy people, 300 mg, 60 days | citrate and an amino acid chelate better than oxide; oxide like placebo | controlled study |
| 7 | Kappeler et al. 2017, BMC Nutr | 20 healthy men, one dose each | citrate: more magnesium in urine and blood than oxide | controlled study, funded by a magnesium supplement manufacturer |
| 8 | Firoz & Graber 2001, Magnes Res | healthy people, 4 products | oxide around 4% absorption; chloride, lactate, aspartate equally good and better | small study |
| 9 | Schutten et al. 2022, J Am Heart Assoc | 164 adults, 450 mg, 24 weeks | blood level equal for citrate, oxide, sulfate; urine: citrate ahead; oxide fewer digestive complaints than citrate; no effect on blood pressure | large controlled study, co-funded by a magnesium raw material manufacturer |
| 10 | Shechter et al. 2012, Magnes Res | 41 healthy people, 1 month each | oxide (520 mg) raised cellular magnesium more than citrate (296 mg) | controlled study, unequal amounts |
| 11 | Ivanovic et al. 2022, BioMetals | 61 young women, 10 days | oxide best; carbonate effervescent tablets showed no rise | small study |
| 12 | Bøhmer et al. 1990, Magnes Trace Elem | 18 female students | no difference between citrate/lactate, hydroxide and chloride | small study |
| 13 | Pajuelo et al. 2024, Nutrients | 40 healthy people, one dose each | no measurable rise after bisglycinate; rise after oxide and citrate | controlled study, manufacturer study of a competing form |
| 14 | Schuette et al. 1994, JPEN | 12 patients after bowel surgery | overall equal; glycinate ahead with poor absorption; better tolerated than oxide | small controlled study |
| 15 | Blancquaert et al. 2019, Nutrients | 15 products in the lab, 30 people | dissolution and absorption vary widely by product | lab + small study |
| 16 | Coudray et al. 2005, Magnes Res | 80 rats, 10 forms | absorption 50–67%, organic forms slightly ahead | animal study |
| 17 | Mori et al. 2021, Nutrients | review | magnesium oxide long used as a laxative in East Asia | review article |
| 18 | Uysal et al. 2019, Biol Trace Elem Res | rats, 5 forms | malate stayed highest longest in blood; acetyltaurate highest in brain | animal study |
| 19 | Russell et al. 1995, J Rheumatol | 24 people with fibromyalgia | magnesium + malic acid: no effect in the blinded phase | small controlled study |
| 20 | Ferreira et al. 2019, Medwave | evaluation of 7 reviews | magnesium + malic acid: little or no difference for pain | evidence review |
| 21 | Slutsky et al. 2010, Neuron | rats | threonate raised brain magnesium, better learning | animal study |
| 22 | Zhang et al. 2022, Nutrients | 109 adults, blend with threonate, 30 days | better memory tests | controlled study, funded by the raw material manufacturer |
| 23 | Hausenblas et al. 2024, Sleep Med X | 80 adults with sleep problems, 1 g, 21 days | better sleep by questionnaire and sleep-tracking ring | controlled study, funded by the raw material distributor |
| 24 | Lopresti & Smith 2026, Front Nutr | 100 adults, 2 g, 6 weeks | better memory tests; sleep only partially better | controlled study, funded by the raw material manufacturer |
| 25 | Gröber et al. 2017, Nutrients | review | absorption through the skin not proven | review article |
| 26 | McFarlin et al. 2025, J Med Food | 20 healthy people | two seawater products: more magnesium in urine than placebo, equal to each other | small study, funded by the manufacturer |
| 27 | Felice et al. 2018, Nutrients | cell study | seawater product similar to chloride, better than oxide | cell study, author employed by the manufacturer |
| 28 | Merschmann et al. 2025, Mol Nutr Food Res | controlled studies in healthy people | citrate and chloride good bioavailability, oxide poor | review article |
| 29 | EFSA / SCF 2001 | official scientific opinion | upper limit of 250 mg/day from supplements | official assessment |
| 30 | Schuster et al. 2025, Nat Sci Sleep | 155 adults, 250 mg bisglycinate, 4 weeks | sleep problems decreased more than with placebo (small effect) | controlled study |
| 31 | Shrivastava et al. 2019, J Tradit Complement Med | rats | magnesium taurate: effects on blood pressure and heart in the animal model | animal study |
| 32 | Rosner et al. 2023, Mayo Clin Proc | review article | build-up with reduced kidney function | review |
| 33 | Argeros et al. 2025, Hypertension | 38 controlled studies, 2,709 participants, at least 4 weeks | blood pressure −2.8/−2.1 mmHg vs placebo; clearer in people on blood pressure medication and with low magnesium levels; not significant at normal blood pressure | meta-analysis, high heterogeneity between studies |
Author: PRONESIUM Editorial Team · Last reviewed: 29/09/2026 · This article does not replace medical advice.
