In short
KSM-66 is a root-only ashwagandha extract standardised to 5% withanolides. Across controlled trials it modestly lowers perceived stress and serum cortisol and improves sleep quality and time to fall asleep, most of that evidence at 600 mg per day, in small studies, several manufacturer-funded. It is not for everyone; see the safety section. We use 300 mg in the evening as one part of a combined recovery layer, not as a standalone stress dose. Root-only sourcing is a quality decision. Inside LAYER THREE it works alongside magnesium bisglycinate, L-theanine, valerian and lemon balm: five routes converging on the same evening downshift.
01What it is
Ashwagandha (Withania somnifera) is a plant used in Ayurvedic practice for centuries. KSM-66 is a specific branded extract made from the root only, standardised to at least 5% withanolides, the group of compounds used as its quality marker. In LAYER THREE it appears at 300 mg per day, which delivers roughly 15 mg of withanolides.
Two details matter before any claim: the plant part (root versus leaf) and the dose. Both are specific decisions, and both are covered below rather than glossed.
02How it works
The most supported mechanism is the stress axis. In human trials, ashwagandha lowers morning and serum cortisol, the primary hormone of the HPA (hypothalamic-pituitary-adrenal) stress response.1 That the cortisol change is what causes the felt effect is a reasonable inference, not a proven chain. A biomarker moving is not the same as a mechanism demonstrated end to end.
A second, more speculative mechanism is often repeated: that ashwagandha acts on GABA, the brain's main calming neurotransmitter. The evidence here is thin and worth stating honestly. A single laboratory study found a root extract activated GABA-type receptors, but the isolated withanolides did not produce the effect; an unidentified constituent did.2 So the calming mechanism is in-vitro only, and the withanolides used to standardise the extract are a quality marker, not the proven active. We flag this rather than present a tidy story.
Moderate · cortisol Mechanistic · GABA
Cortisol lowering is a real, repeated biomarker finding in humans. The GABAergic and sedative mechanism rests on one in-vitro study and is not attributable to withanolides. We do not state it as fact.
03What the evidence shows
Stress and cortisol
The best-known trial gave 600 mg per day of KSM-66 for 60 days to 64 chronically stressed adults and reported substantially lower serum cortisol and stress scores against placebo.3 A later study found reductions at both 250 mg and 600 mg per day, with the larger dose producing the larger effect.4 The direction is consistent. The caveats are consistent too: small samples, mostly single-site, several run by author groups affiliated with the extract's maker.
Newer syntheses hold the same line. A 2024 systematic review and meta-analysis confirmed reductions in stress and anxiety scores against placebo.9 A 2025 meta-analysis found cortisol falls under supplementation even where perceived stress does not always move, which is the honest shape of the evidence: a clear hormone level signal, a more modest experience level one.10
Sleep
In adults with insomnia and anxiety, 600 mg per day for ten weeks shortened the time taken to fall asleep and improved sleep efficiency and quality on actigraphy.5 An independent meta-analysis of five trials confirmed a real but small effect on sleep, most pronounced in people with diagnosed insomnia and at 600 mg per day for eight weeks or more.6 This is the honest shape of the evidence: a genuine, modest effect, clearest at a higher dose than ours and in worse sleepers than the average person.
04Why root-only
KSM-66 is made from the root and nothing else. The leaf and aerial parts of the plant concentrate withaferin A, a withanolide that is potently cytotoxic in laboratory settings and is not something you want in a daily supplement. India's own food-safety framework restricts leaf use, and much of the regulatory scrutiny in Europe centres on leaf-containing or mislabelled products.7
We choose root-only sourcing as a quality-control and safety discipline. To be precise about the claim: there is no head-to-head human trial proving root is safer than leaf. The substantive point is verifiable sourcing of a well-characterised plant part, not a proven clinical advantage. We would rather state that exactly than imply more.
05Why 300 mg, not 600
Most of the headline results above used 600 mg per day. We use 300 mg. We will not pretend those are the same, so here is the reasoning.
LAYER THREE is not a stress megadose in a single capsule. It is a combined evening layer, with magnesium bisglycinate, L-theanine, valerian, lemon balm and B6 alongside the ashwagandha, where several components support the same evening downshift through different routes. In that context 300 mg contributes to the layer without carrying it alone, and keeps a comfortable margin below doses associated with more frequent side effects. It also keeps the daily withanolide load at a level that respects the stricter national limits in parts of the EU. If you are looking specifically for the 600 mg single-ingredient stress protocol from the trials, this is deliberately not that. It is the evening layer's version of the same evidence: the studied extract, the studied direction, sized for every night.
06Why the evening
Ashwagandha's most relevant effects here, lower cortisol and easier sleep onset, belong to the end of the day, not the start. Placing it in the evening layer aligns the compound with the physiology it supports: the downshift out of a working day and into recovery. The morning and daytime layers are built around different compounds for different jobs.
07What it likely does not do
- Boost testosterone. The between-group signal is weak or null and comes largely from a different extract. We keep it out of the evening-recovery story.
- Improve general quality of life. The sleep meta-analysis found no significant effect on that broader outcome.6
- Sharpen focus. Cognitive trials use different doses and timing; borrowing them for an evening dose would be dishonest.
- Match the 600 mg trial results at 300 mg. Stated plainly so it is not implied; what 300 mg does here is carry ashwagandha's share of a five ingredient downshift, at a dose with a wide comfort margin.
08The research horizon
Ashwagandha has entered the phase where the literature consolidates rather than accumulates. A 2026 meta-analysis of 20 randomised trials and 1,249 participants found improvements across memory, attention, and physical performance measures.11 A second 2026 meta-analysis confirmed reductions in stress and anxiety and found something rarer: a significant dose-response relationship, the pattern pharmacology trusts most, because an effect that scales with dose is an effect that is real.12 New trials keep arriving on top: an eight week randomised trial reported improvements in working memory, attention accuracy, and mood,13 and the frontier now includes novel delivery forms tested in controlled conditions14 and registered trials running through 2026 in new populations. Few botanicals attract this volume of study, and the direction has held through all of it. This page moves with the literature.
09Safety and who should avoid it
Please read this part
Ashwagandha is generally well tolerated in short-term use, but it is not for everyone. Do not take it if you are pregnant or breastfeeding, or under 18. Speak to a doctor first if you have a thyroid condition or liver disease, or if you take sedatives or other medication. Rare cases of reversible liver injury have been reported: a handful of documented cases worldwide against millions of daily users, all in the published series resolving after stopping. The full picture is on the safety page.8 Regulatory status varies across the EU and is under review in several countries. The discussion concerns high dose, long term use patterns; the 300 mg root only extract here sits at the conservative end of the studied range. Stop use and seek advice if you notice symptoms such as unusual tiredness, yellowing of the skin or eyes, or itching. This is educational information, not medical advice.
We treat this warning as part of the product, not a disclaimer bolted on. A supplement that tells you who should not take it is easier to trust than one that does not.
Sources
Next in this thread
Why recovery starts before bedThe physiology the evening layer is built around, and what actually drives the downshift.→How we handle evidence. Claims are graded and linked to primary sources; industry funding and study size are disclosed. This page describes what research shows about an ingredient. It is not a health claim about any product and not medical advice. Food supplements do not replace a varied diet or a healthy lifestyle. Do not exceed the recommended daily dose.