Ingredients

Standardized Extracts: What the Percentage on the Label Means

Published July 12, 2026 · The Ingredient Brief

Two bottles can name the same plant, declare the same number of milligrams, and deliver amounts of the measured compound that differ by a factor of twenty. Neither label is lying. The information that separates them is printed on both — in the part everyone skims.

What standardization actually specifies

A plant is not a compound. It is a few hundred compounds in proportions that shift with the soil, the season, the harvest, and the drying shed. Standardization is the manufacturing step that fixes one of those proportions: the extract is processed and adjusted so that a named constituent — the marker — arrives at a declared percentage of the extract's weight, batch after batch.

That is what “standardized to 95% curcuminoids” asserts. Not that the product is potent, not that it is good, not that it will do anything. It asserts a measurable, testable quantity: in every 100 mg of that extract, roughly 95 mg is the family of compounds called curcuminoids. It turns a botanical into something with a specification — the precondition for comparing two products at all.

Two limits belong in the same breath. First, the marker is not necessarily the thing that matters. As the NIH Office of Dietary Supplements puts it, manufacturers identify constituents called markers and adjust them to keep batches consistent — but for most botanicals the constituents actually responsible for the plant's effects are not known, so markers often serve as surrogates. A standardized extract is a consistent extract, and consistency is not efficacy.

Second, there is no US legal definition of the word. Federal law does not require supplements to be standardized, and there is no legal or regulatory definition of “standardized” in the United States. The percentage is a claim the manufacturer makes about its own material, not a term of art the government polices. That does not make it worthless — it makes it a checkable claim rather than a guaranteed one, and checkable is the whole game.

Why two honest labels can be twenty times apart

Take the arithmetic straight off two hypothetical panels, both naming turmeric root, both declaring 500 mg.

Same plant, same milligrams

What the panel saysProduct AProduct B
IngredientTurmeric root powderTurmeric root extract
Amount per serving500 mg500 mg
Standardization3% curcuminoids95% curcuminoids
Curcuminoids delivered15 mg475 mg

Same plant. Same part of the plant. Same headline number. A thirty-one-fold difference in the compound the label chose to measure — and both statements are true. The front of the bottle is not where the difference lives. The percentage is.

Notice what the arithmetic does not establish: that Product B is the better purchase. Whole-plant powder and concentrated extract are not two grades of the same thing — they are different materials. The powder carries the plant's full matrix at its natural concentration; the extract carries a selected fraction of it, multiplied. Research is run on one specific preparation at one specific dose, and the findings travel with that preparation. If the evidence you are leaning on used a 95% extract, a 3% powder is not a cheaper version of it. It is a different input, and the arithmetic above tells you by how much.

Ratio extracts: 10:1 of what, measured how?

The other number you will see is an extraction ratio — 4:1, 10:1, sometimes 50:1. It is meant to say that ten kilograms of raw plant material were reduced to one kilogram of finished extract.

Read closely and the ratio is a statement about a process, not about a content. It answers "how much starting material went in." It does not answer the only question that lets you compare anything: how much of the compound you care about comes out. Three gaps open up immediately:

  • Ten to one of what? Ten kilos of leaf and ten kilos of root are not the same ten kilos. Neither are a rich harvest and a depleted one. The ratio inherits every bit of variability the raw material had.
  • Concentrated by which measure? A 10:1 reduction by total weight does not mean the marker was concentrated ten-fold. Extraction is selective — the solvent pulls some compounds hard and leaves others behind. Weight ratio and marker ratio are different numbers.
  • Ratio to which finished powder? Extracts are often blended with carriers to make them handleable, and a ratio calculated before the carrier describes a material that is not what is in the capsule.

A ratio with no named marker and no percentage is therefore not a potency statement. It is a production note. Put the sharpest version on a card: a percentage tells you what came out; a ratio only tells you what went in. When a label gives the ratio alone, the honest conclusion is not "this is weak" but "potency is not stated" — a different and more useful verdict.

The part of the plant is on the label, and it is load-bearing

US regulation is unusually specific here, and in your favor. Under 21 CFR 101.4, a botanical ingredient must be followed by a statement of the part of the plant from which the dietary ingredient is derived — the rule's own examples are "Garlic bulb" or "Garlic (bulb)" — plus the Latin binomial name. The Supplement Facts rules at 21 CFR 101.36 carry the same requirement into the panel itself: botanicals must specify the plant part, whether they appear in the ingredient statement or in the nutrition label.

This matters because root, leaf, aerial part, fruit, bark and seed are chemically distinct raw materials sharing one common name. Ginkgo leaf and ginkgo seed are not interchangeable inputs. Ashwagandha root and ashwagandha leaf yield different constituent profiles. The marketing on the front of the box says the plant; the regulation forces the panel to say the part. Treat it as a gate: if two products differ there, you are not comparing two doses of one thing, and no percentage will fix that.

The tool: comparing two extracts of the same plant in four steps

Two minutes, two labels. It replaces every impression the front of the package is working to give you with one number you computed yourself.

  1. Find the marker. On each panel, locate the named constituent the extract is standardized to — curcuminoids, silymarin, flavone glycosides, withanolides. Confirm both products name the same marker and the same plant part. If a label names no marker, stop: that product has not stated its potency, and no arithmetic can recover it.
  2. Find the percentage. Read the figure attached to that marker — "standardized to 24% flavone glycosides," "min. 80% silymarin." One extract may declare two markers at two percentages; if so, run the math on both.
  3. Multiply by the dose of the extract. Take the milligrams of extract per serving (not the total capsule weight, which includes carriers) and multiply by the percentage. 300 mg at 24% is 72 mg of marker. 500 mg at 95% is 475 mg. If the panel gives servings of two capsules, do the math per serving, not per capsule — that is the most common place people lose a factor of two.
  4. Compare the results, not the headlines. You now hold two numbers in the same unit, for the same compound, from the same plant part. That is the only comparison the labels support — and a 500 mg product delivering 15 mg of marker is nowhere near a 300 mg product delivering 72 mg of it, whatever the front of the bottle implies.

One honest caveat, and it is why this page will not tell you that bigger is better: the number you just computed is a potency figure, not a verdict. It tells you what you are getting, in the units the research uses. What that amount is worth is a separate question, and it belongs to the evidence, not to the label. The four steps exist to stop you comparing two products on a dimension neither one measured — and to hand back the one they both did.

The short version

  • A standardization percentage fixes how much of a named marker is in the extract. It buys you consistency and comparability — not proof that the marker is what matters.
  • A ratio (10:1) describes what went in; a percentage describes what came out. A ratio with no marker and no percentage is not a potency claim.
  • The plant part is required on the label (21 CFR 101.4). Root, leaf and fruit are different raw materials wearing the same common name — check it before comparing anything.
  • Run the four steps: marker → percentage → multiply by the extract dose → compare the milligrams. Two labels that look identical usually stop looking identical at step three.

Sources

  • NIH Office of Dietary Supplements — Botanical Dietary Supplements: Background Information: marker compounds used as surrogates for unknown active constituents; US law does not require standardization and provides no legal or regulatory definition of the term.
  • Code of Federal Regulations — 21 CFR 101.4, Designation of ingredients: botanical ingredients must state the part of the plant from which the ingredient is derived, plus the Latin binomial name.
  • Code of Federal Regulations — 21 CFR 101.36, Nutrition labeling of dietary supplements: source ingredients are identified per § 101.4, and the listing of botanicals must specify the plant part.
  • US Food and Drug Administration — Dietary Supplement Labeling Guide, Chapter IV (Nutrition Labeling; botanicals).