Valerian Has a 40-Year Head Start — and the Least Readable Labels We’ve Counted

Valerian root is the elder of the sleep aisle. Its first placebo-controlled trial ran in 1982 (PMID 7122669); a meta-analysis of sixteen trials existed by 2006 (PMID 17145239). Four decades of research head start on almost everything else on the shelf. So its labels should be the most readable ones, right?

We read 576 valerian-containing products from the NIH label database. They are the least readable labels we have counted in this entire series — and in one specific way, valerian broke our method itself.

Worst in series, on every measure we track

• 44.1% hide the valerian dose in a blend — the highest hidden-dose rate of the eight ingredients we’ve counted. (The category average across sleep formulas was 38.6%; theanine, the best behaved, hides 10.8%.)
• 11.3% state any standardization (valerenic acids). Ashwagandha managed 29%, saffron 30%. Valerian — the herb famous for wildly variable extract quality — discloses the least.
• 54% don’t say “root” anywhere on the valerian line, though the root is what the trials used.
• The dilution effect, eighth replication: median 500 mg in dedicated valerian products, 150 mg once it’s an ingredient in someone else’s formula.

And then the units broke

In every previous study we could at least ask: does the stated dose match what a trial used? Valerian is where that question stops being answerable — and we’d rather show you why than pretend to score it.

Here is a real pattern from our sample, national drugstore brands included: the front of the bottle says “Valerian 450 mg”; the supplement facts panel says 2,250 mg. Neither number is a lie. The 450 is the weight of a concentrated 5:1 extract; the 2,250 is the raw-herb weight it was made from, and the panel is allowed to state the equivalence. Among number-titled valerian products, 16% carry a panel figure four to ten times the number on the front. Fifteen percent of all labels state some N:1 extract ratio; most state none.

Now try to compare any of that to the research. The 1982 trial used 400 mg of an aqueous extract. A modern “450 mg” of 5:1 ethanolic extract is not more, not less — it’s incommensurable. The 2006 meta-analysis said it directly: doses and preparations “varied considerably” across trials. Forty years on, the shelf has faithfully reproduced that chaos, one label at a time.

So we’re not scoring valerian doses against a threshold, and we don’t think anyone honestly can. For seven ingredients we could count. Valerian is where counting itself breaks — which tells you more about the category than any percentage would.

Update, same day: fair pushback — refusing to score is also a choice, so here is the number with its honesty intact. Scored exactly as printed, 45% of stated doses reach the 400 mg of the 1982 trial. Normalized to extract weight — dividing herb-equivalents by their stated N:1 ratios and preferring the front-label extract figure where the panel is its 4–10× multiple — the median drops to 250 mg and the share reaching 400 falls to 38%. So: between 38% and 45%, depending entirely on what you decide the milligrams mean — a seven-point swing produced by unit ambiguity alone, which rather proves the point of this section. The aqueous-vs-ethanolic caveat still stands on top of both numbers.

Where we stand

Valerian is on our published list of rejected ingredients — we passed on it for inconsistent trial results and exactly this extract-quality chaos, before we ran this count. So we’re not neutral: this data flatters our decision. Weigh accordingly; the dataset is public.

Limits: our sample is the 600 most relevant of 948 database records (578 retrieved, 576 with valerian in the panel). Medians are “as printed” — given the unit chaos above, treat the dedicated-vs-formula comparison as a pattern, not a precise ratio. Root and standardization detection read the label text; a bottle that used root but didn’t say so counts as silent, because silence is what the buyer sees.

How to read a valerian label in ten seconds

1. Find out what the milligrams are. Extract weight or herb equivalent? If the label doesn’t say — and most don’t — the number can’t be compared to anything.
2. Look for “root” and a valerenic-acid percentage. One label in nine has the second one. That’s the label taking its own herb seriously.
3. In a blend, assume the dose answers to the blend’s marketing, not to valerian’s trials. 44% of the time you won’t see the dose at all.

How we counted, so you can check us

We searched the DSLD for “valerian”, took the 600 most relevant on-market records, and read 578 full labels on 30 August 2026; 576 list valerian in the supplement facts panel. Doses were read from the valerian line (blends walked, units converted to milligrams, per serving as printed). Front-label numbers were compared to panel figures on the 75 products carrying a milligram figure in the product name. Both cited trials are linked above; the 1982 trial’s 400 mg aqueous-extract dose and the meta-analysis’s “varied considerably” are taken from their abstracts.

The database is free and needs no account. If you repeat this and get different numbers, we’d like to know.

†These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease. This article is educational and is not medical advice; if you have a medical condition or take medication, talk to your healthcare provider.