Comparison guide
Which magnesium should you take?
Magnesium forms differ in two things that matter: how much elemental magnesium you absorb, and what the carrier molecule does. The rest is marketing. Whichever form you choose, the target is 200–400 mg of elemental magnesium — check the label, because compound weight is not magnesium weight.
The options, compared
| Option | Evidence | Best for | The honest note | |
|---|---|---|---|---|
| Magnesium | Moderate evidence | Most people — evenings, sleep, tolerance | The glycinate/bisglycinate form: well absorbed, gentle on the gut, and the glycine adds a mild calming effect. | View |
| Magnesium Citrate | Moderate evidence | Best value + constipation | Good absorption at a lower price; the mild laxative effect is a feature if you're constipated, a bug if you're not. | View |
| Magnesium Malate | Limited evidence | Daytime use | Good absorption without sedation; popular for energy claims that remain unproven. | View |
| Magnesium Taurate | Limited evidence | Blood-pressure focus | Pairs magnesium with taurine, which has its own modest BP data — direct comparative trials are scarce. | View |
| Magnesium L-Threonate | Limited evidence | Brain claims — premium price, early data | The only form with brain-penetration data, but human trials are small and each capsule carries little elemental magnesium. | View |
| Magnesium Orotate | Limited evidence | Niche cardiac use | Some heart-failure adjunct data; expensive per mg of magnesium. | View |
| Magnesium Chloride | Moderate evidence | Topical/liquid preference | Decent oral absorption; 'magnesium oil' sprays lack good evidence for transdermal delivery. | View |
| Magnesium Oxide | Moderate evidence | Laxative use only | Cheapest and most common in multis, but poorly absorbed — fine as a laxative, poor as repletion. | View |
Our verdict
Glycinate at night if your goal is sleep, stress, or gut comfort; citrate if you want to spend less or also need help with regularity. Skip oxide except as a laxative, and treat threonate as an experiment, not a proven upgrade.