Ingredient deep-dive

Ashwagandha: The Dose Its Own Trials Actually Used

Ashwagandha is the best-studied botanical in the calm-and-sleep aisle, and its trial literature is unusually specific about dose. That makes it the single most useful ingredient for judging a formula, because the number a label would have to reach is published, repeated and easy to check.

By Quiet Tone Team · Last updated October 29, 2026 · 8 min read · Ingredient deep-dive · Reviewed by Dr. Kelsee Lothes, Au.D

What ashwagandha is, briefly

Withania somnifera, a shrub grown across India and parts of Africa, used in Ayurvedic practice for a very long time and studied in controlled trials for about twenty years. The root is the part with the human literature behind it.

It is usually described as an adaptogen, a word that means something more specific in marketing than it does in pharmacology. The useful version is narrower: in controlled trials it has been studied against measures of stress, sleep and related outcomes, at doses that are published and consistent.

Why the plant part is printed on better labels

Root and leaf are different materials. Nearly all of the human trial literature used root or root extract, and a label that specifies the root is telling you it matches the material the research used.

A label that says only Ashwagandha is leaving the question open, and the difference is not trivial: much of the variability between botanical products comes from a supplier using a cheaper part of the same plant.

The doses the trials actually ran

This is the section worth photographing. The trial doses are specific and they repeat.

Doses read from each paper's own abstract. None of these trials studied a multi-ingredient formula.

The safety and tolerability trial in 1,002 adults is the largest of them and used 600 mg a day in two divided doses over eight weeks. A three-arm controlled trial used the same total as 300 mg twice a day. A dose-response meta-analysis pools the published range and describes where the curve sits.

Six hundred milligrams a day. That is the number, it is remarkably consistent across the literature, and it is what a label would have to reach to be doing what these trials did.

Why the split dose keeps recurring

Almost every trial in this list uses two doses a day rather than one, and that is a design choice worth understanding rather than a detail to skip.

Splitting a botanical dose spreads exposure through the day and reduces the peak, which generally improves tolerability. It also matches how the traditional preparations were taken. Whatever the reason, it is what the trials did, and a once-daily product is not reproducing that schedule even if it reached the total.

What that means for a once-a-day capsule

It means two departures from the trial design rather than one: a different total and a different schedule. Neither is fatal on its own, and stating both is more useful than pretending a single capsule reproduces a twice-daily regimen.

It also means a product that splits its own dose across two capsules is closer to the research than one that does not, which is a small point in favour of two-capsule formats that rarely gets made.

Extract ratios, and why a milligram is not a milligram

One further complication, and it is the reason two labels both printing 600 mg can be carrying different things.

Botanical ingredients are usually extracts, and an extract carries a ratio or a standardisation. A 10:1 extract means ten parts of raw root reduced to one part of extract. A standardised extract is specified to contain a stated percentage of a marker compound. Those are different claims, and a bare milligram figure tells you neither.

• A raw powder at 600 mg and a concentrated extract at 600 mg are not the same material.
• A standardisation percentage tells you what fraction of the weight is the marker compound being counted.
• Trials generally specify the extract they used, which is why the paper is a better source than a summary of it.
• A label printing a plain milligram figure with no ratio and no standardisation is giving you the least informative version of the number.
• None of this is a reason to dismiss a figure; it is a reason to read the whole row rather than the number alone.

The liver reports, in proportion

Any honest article about this plant has to cover the safety signal, and covering it proportionately is harder than covering it at all.

Rarity matters here and so does dose. A signal recorded at a therapeutic single-ingredient dose does not automatically transfer to a fraction of a blend.

A case series from the US drug-induced liver injury network and a series from India both record liver injury associated with ashwagandha, reversible in the reported cases. NCCIH carries the plain-language summary, including how rare it is against how widely the ingredient is used.

Two things belong beside those reports. The first is that they involve single-ingredient products at therapeutic doses, which is not the same exposure as a share of a small blend. The second is that a rare signal in a very widely used ingredient is still a signal worth knowing about, particularly for anyone with existing liver disease.

What to watch for

Yellowing of the skin or the eyes, dark urine that is not the harmless bright yellow a B-complex produces, persistent pain under the right ribs, or unexplained fatigue starting after a new supplement. Any of those is a reason to stop and be seen rather than to wait and see.

What this means on a multi-ingredient panel

Put the 600 mg figure beside a typical hearing or sleep formula and the arithmetic is usually decisive in one direction.

If the panel quantifies ashwagandha, compare directly. If the panel puts it in a blend, compare 600 mg against the blend total. A blend smaller than 600 mg cannot contain 600 mg of any one member, and no further information is needed to establish that.

1. Find the ashwagandha row and note whether it carries a figure.
2. If it does, compare it with 600 mg and note the ratio.
3. If it does not, find the blend total and compare that with 600 mg instead.
4. Check whether the plant part is specified as root.
5. Check whether an extract ratio or a standardisation is printed anywhere on the row.

Five steps, and most labels in this category fail at step three. That is not a reason to avoid them; it is a reason to know what is being bought, which is a wide formula rather than a therapeutic dose of one plant.

Who should check before taking it

• Anyone with existing liver disease, because of the case series above.
• Anyone taking thyroid medication, since ashwagandha has been reported to affect thyroid measures.
• Anyone pregnant or nursing, for whom the safety data does not exist.
• Anyone taking sedatives, since the expected direction is additive.
• Anyone about to have surgery, on the general principle of pausing supplements beforehand.

That list is short and specific on purpose. A general caution attached to every ingredient teaches a reader to skip cautions, and the five groups above have an actual reason to pause.

Four questions for any ashwagandha row

1. How many milligrams, and is it quantified at all? 600 mg a day is what the trials used.
2. Is the root specified? Nearly all the human evidence is root.
3. Is an extract ratio or standardisation printed? Without one, a milligram figure is only half a specification.
4. Is the dose split, as the trials split it? A once-daily product is a different schedule from the published one.

Four questions, and a label answering all four is doing something unusual for this category. Most answer one or two, which is a useful thing to establish in the thirty seconds it takes.

References