Biohacking
Natural Sleep Aids, Graded by Evidence (Not by Marketing)
Magnesium, glycine, apigenin, and the rest of the natural sleep-aid stack, ranked by RCT evidence and real effect size, not marketing claims.
Every "natural sleep aid" list ranks the same dozen ingredients by how well they sell, not by what a randomized trial found. Magnesium, glycine, apigenin, valerian, GABA, and tart cherry all show up in nearly identical roundups, usually with the same three vague adjectives: "calming," "supportive," "gentle." None of that tells you which one has an actual dose-response trial behind it and which one has never been tested alone in a human. This post grades each ingredient by the strength of its evidence, not its shelf presence, as part of the biohacking framework I use to decide what's worth paying for.
Medical disclaimer. I'm not a physician. This is personal research and general information, not medical advice. Several ingredients below interact with prescription sedatives, blood thinners, antidepressants, and blood pressure medication. Talk to a doctor before starting any sleep supplement, especially if you take other medication or have a diagnosed sleep or mood disorder.
Bottom Line. (1) Magnesium bisglycinate, L-theanine, ashwagandha, and melatonin have real randomized controlled trials behind them, though effect sizes are mostly small to moderate. (2) Melatonin's efficacy evidence is strong, but the products themselves are not: a 2023 JAMA analysis found 88% of gummies mislabeled by more than 10%. (3) Glycine, whole chamomile extract, and tart cherry juice are plausible but under-studied, with small trials and real mechanistic uncertainty. (4) Isolated apigenin capsules, valerian, GABA supplements, and "serotonin foods" are mostly marketing built on mechanisms that don't hold up in whole humans. (5) None of this replaces the free interventions in sleep optimization — supplements are the last 10%, not the first move.
In this article:
- How I Grade "Natural Sleep Aid" Evidence
- The Aids With Real Trial Support
- The Aids That Are Plausible but Under-Studied
- The Aids That Are Mostly Marketing
- Dosing, Timing, and What to Combine (or Not)
- Frequently Asked Questions
How I Grade "Natural Sleep Aid" Evidence
I sort each ingredient into one of three tiers, the same way I'd grade a trading system before risking capital on it: has it been tested against a real comparison, and does the effect size survive contact with a larger sample?
- Real trial support. Multiple randomized, placebo-controlled trials, ideally pooled in a meta-analysis, with a stated effect size (not just "participants reported improvement").
- Plausible but under-studied. At least one credible trial or a mechanistic basis worth taking seriously, but small samples, high risk of bias, or results that haven't been replicated at scale.
- Mostly marketing. No trial on the isolated ingredient at the marketed dose, a mechanism that doesn't survive scrutiny (it can't cross the barrier it needs to cross, or the food-level dose is too small to matter), or repeated null results in the most rigorous available trials.
Anecdote and testimonial content, however common in "best natural sleep aid" roundups, doesn't move an ingredient between tiers. A supplement brand's own funded study doesn't either, unless it's been replicated independently.
The Aids With Real Trial Support
Magnesium Bisglycinate
The largest placebo-controlled trial on magnesium and sleep to date, published in 2025, randomized 155 adults with self-reported poor sleep to 250 mg of elemental magnesium (as bisglycinate) or placebo for four weeks. The magnesium group's Insomnia Severity Index score dropped by 3.9 points versus 2.3 points for placebo, a statistically significant difference (p = 0.049). The effect size was small (Cohen's d = 0.2) (Nature and Science of Sleep, 2025, retrieved 2026-09-06). The benefit was largest in people who already reported low dietary magnesium intake — this looks like it's correcting a deficiency more than adding a sedative effect on top of an adequate baseline.
Dose in the trial: 250 mg elemental magnesium daily, as bisglycinate. For the broader case for this form over cheaper oxide, see magnesium supplements for energy.
L-Theanine
A 2025 meta-analysis pooled 19 randomized controlled trials with 897 participants, at doses of 200-450 mg/day (Sleep Medicine Reviews, Bulman et al., 2025, retrieved 2026-09-06). It found significant improvement in subjective sleep onset latency in 10 of 19 studies, overall sleep quality score in 12 of 19, and daytime dysfunction in 9 of 19. It doesn't sedate the way melatonin or a benzodiazepine does; the mechanism is a mild anxiolytic effect through alpha-wave activity, which is why it shows up more consistently for sleep onset than for total sleep time.
Dose in the trials: 200-450 mg, taken in the evening.
Melatonin
Melatonin has the deepest efficacy record on this list. A 2024 dose-response meta-analysis pooled 26 randomized controlled trials, covering 1,689 observations from 1987 to 2020 (Journal of Pineal Research, 2024, retrieved 2026-09-06). It found effects on sleep onset latency and total sleep time increased with dose up to 4 mg/day, with 2 mg/day already significantly outperforming placebo and 1 mg. The efficacy evidence is real. The product-quality evidence is not: a 2023 JAMA analysis of 25 melatonin gummies found 22 (88%) contained melatonin outside ±10% of the labeled amount, ranging from 74% to 347% of the stated dose, and one product had no melatonin at all (JAMA, Cohen et al., 2023, retrieved 2026-09-06).
Practical takeaway: the dose on the label of a gummy is closer to a suggestion than a fact. A capsule or liquid from a manufacturer that publishes third-party lab testing is a safer bet than a gummy.
Ashwagandha
A 2021 meta-analysis of 5 randomized controlled trials (400 participants) found ashwagandha extract had a small but significant effect on overall sleep (standardized mean difference 0.59), with larger effects in adults with diagnosed insomnia (0.84) versus healthy adults (0.63), and at doses of 600 mg/day or more taken for 8 weeks or longer (PLOS ONE, 2021, retrieved 2026-09-06). This is the one entry on this list where patience matters more than the pill: the trials that found an effect ran 8 weeks minimum, not 8 days.
Dose in the trials: ≥600 mg/day root extract, for at least 8 weeks.
The Aids That Are Plausible but Under-Studied
Glycine. A January 2024 systematic review of glycine's effects on human health concluded that small crossover studies consistently report improved subjective sleep quality and shorter time to both sleep onset and slow-wave sleep after 3 g of glycine before bed — but the studies were small and at high risk of bias (Nutrients, 2024, retrieved 2026-09-06). The direction of the effect is consistent across every trial that's run it; the sample sizes just haven't caught up to a firm conclusion.
Chamomile (and the apigenin confusion). A 2024 meta-analysis of 10 trials (772 participants) found chamomile extract significantly reduced Pittsburgh Sleep Quality Index scores, with the clearest benefit on nighttime awakenings rather than total sleep duration (Complementary Therapies in Medicine, Kazemi et al., 2024, retrieved 2026-09-06). The catch: every one of those trials tested whole chamomile extract, a mixture of dozens of compounds, not isolated apigenin. No published human trial has tested apigenin alone at the doses sold in supplement capsules — chamomile's track record and an apigenin pill's track record are not the same evidence.
Tart cherry juice. A 2025 systematic review of 7 interventional studies found mixed results: some reported longer sleep duration and higher melatonin levels after tart cherry consumption, others didn't (Food Science & Nutrition, 2025, retrieved 2026-09-06). The melatonin content itself is tiny — about 42.6 micrograms per 30 mL of concentrate, a fraction of even a low-dose supplement — so if there's a real effect, it's more likely from procyanidin B-2 sparing tryptophan than from the melatonin content people assume is doing the work.
Personal data — pending measurement. This section will carry a 4-week self-tracked comparison: magnesium bisglycinate (250 mg) versus a 3 g glycine dose, each run for 2 weeks with a 1-week washout, logged against sleep-onset time and morning grogginess on a 1-5 scale. Not filled in yet: I publish it once the 4-week log is complete, not before.
The Aids That Are Mostly Marketing
Isolated apigenin capsules. As noted above, chamomile has real trial support; a standalone 50 mg apigenin capsule does not. Marketing pages borrow chamomile's evidence and apply it to a different, untested product.
Valerian root. A 2020 systematic review and meta-analysis pooling 60 studies (6,894 participants) found the most recent, most methodologically rigorous trials showed no significant difference from placebo for subjective sleep quality, even though valerian carries a strong safety record (Journal of Evidence-Based Integrative Medicine, Shinjyo et al., 2020, retrieved 2026-09-06). Decades of use and a large trial base haven't produced a consistent, replicated effect.
GABA supplements. GABA is the brain's primary calming neurotransmitter, which is exactly why the marketing pitch is intuitive and exactly why it likely doesn't work orally: whether oral GABA meaningfully crosses the blood-brain barrier is contested in the research, and a 2020 systematic review found only limited-to-moderate evidence for stress and very limited evidence for sleep (Frontiers in Neuroscience, 2020, retrieved 2026-09-06). Taking a molecule your brain already makes doesn't help if it can't get to where it needs to act.
"Serotonin foods for sleep." This is the most mechanistically broken claim on the list. Turkey, bananas, and other tryptophan-containing foods are marketed as serotonin (and by extension melatonin) boosters. Purified tryptophan does raise brain serotonin — but tryptophan competes with every other large amino acid in a normal meal for the same transporter across the blood-brain barrier, so a protein-containing food doesn't reliably raise brain tryptophan the way an isolated dose does. A carbohydrate-heavy, protein-light meal can help indirectly by lowering competing amino acids, but that's a very different claim than "eat turkey to sleep better," and it's not one any bottle on a supplement shelf is making accurately.
Dosing, Timing, and What to Combine (or Not)
- Magnesium bisglycinate: 250 mg elemental magnesium, 1-2 hours before bed. Safe to combine with the others below; watch for loose stools at higher doses.
- L-theanine: 200-450 mg, evening. No known sedative interaction with magnesium or melatonin at these doses.
- Melatonin: Start at 0.5-1 mg, not the 5-10 mg sold in most gummies — higher doses haven't shown proportionally better results and increase next-day grogginess risk. Buy from a brand with third-party testing, not a gummy with no certificate of analysis.
- Ashwagandha: 600 mg/day root extract, taken consistently for at least 8 weeks before judging it. Don't stack with sedating prescription medication without asking a doctor first — it can compound drowsiness.
- Glycine: 3 g before bed, the dose used in every trial so far. It's the best-known research dose, not a proven optimal one.
- What to skip spending on: isolated apigenin capsules, valerian, GABA, and anything marketed specifically as a "serotonin food" supplement.
- What actually needs to happen first: none of this fixes a bad sleep schedule. If the real problem is timing, light, or a phone in bed, the fix is in how to sleep better at night naturally, not a fifth ingredient in this stack.
Frequently Asked Questions
What is the strongest natural sleep aid?
By trial evidence, melatonin has the deepest efficacy record, with a clear dose-response relationship up to 4 mg/day (Journal of Pineal Research, 2024, retrieved 2026-09-06). But "strongest evidence for efficacy" and "safest bet on a store shelf" aren't the same thing, given how often melatonin gummies are mislabeled. Magnesium bisglycinate and ashwagandha have smaller but more reliably dosed effects.
What supplements actually increase deep sleep?
Most of the trials above measured total sleep time, sleep onset, or subjective quality rather than slow-wave (deep) sleep specifically. Glycine is the one exception with direct evidence: small trials report shortened latency to slow-wave sleep specifically, not just total sleep time (Nutrients, 2024, retrieved 2026-09-06). Treat any product claiming to "boost deep sleep" without citing a slow-wave-sleep measurement as marketing language, not evidence.
Do serotonin foods help you sleep?
Not the way they're marketed. Whole foods containing tryptophan don't reliably raise brain tryptophan or serotonin because tryptophan competes with other amino acids for transport across the blood-brain barrier. A carbohydrate-heavy, protein-light evening meal is a more defensible mechanism than any specific "serotonin food."
Grade each ingredient the way a trial would, and the sleep-aid shelf splits cleanly: magnesium, L-theanine, ashwagandha, and melatonin have real trials behind them; glycine and chamomile are promising but thin; valerian, GABA, isolated apigenin, and serotonin-food claims are mostly marketing wearing a lab coat. Start with the free structural fixes, add one supplement at a time, and give each one the 4-8 weeks its own trial data used before deciding it doesn't work.
About this guide. Written by Nate Harmon, an operator and self-experimenter documenting personal protocols in biology for Peak Human Ops. Not a physician or a licensed adviser. Sources are tier 1-3 peer-reviewed studies and government research letters, each linked inline with a retrieval date. How this site is written and corrected is set out in the editorial policy and the corrections log. Who is behind it is on the about page, and you can contact me directly.