Proprietary Blends: What a Supplement Label Legally Has to Tell You, and What It Can Hide

The Supplement Facts panel is a floor, not a disclosure

Most people read a Supplement Facts panel as if it were a full account of what is in the bottle. It is not. It is the minimum a manufacturer is legally obliged to print. Everything above that minimum is voluntary, and the gap between the two is where most of the marketing lives.

In the United States, dietary supplement labelling is governed by 21 CFR 101.36. That regulation sets out which nutrients need a percent Daily Value, how the panel is formatted, and how "other dietary ingredients" — the botanicals, amino acids and extracts that make up most nootropic formulas — must be declared. Botanicals have no established Daily Value, so what you get for them is a name and, if the brand chooses, a number.

The proprietary blend provision sits at paragraph (c). It says a blend of dietary ingredients is identified by the term "Proprietary Blend" or another descriptive or fanciful name, that the ingredients inside it are declared "in descending order of predominance by weight" (101.36(c)(2)), and that "the quantitative amount by weight specified for the proprietary blend shall be the total weight of all other dietary ingredients contained in the proprietary blend" (101.36(c)(3)). Read that second clause carefully. The quantity that must appear is the total. Individual amounts are not mentioned, because they are not required.

This is not a loophole in the ordinary sense. It was written deliberately, so that a company with a genuinely novel formulation ratio could protect it from being copied off the back of the bottle. That rationale is real. It is also, in practice, applied to formulas whose composition is entirely obvious and whose only secret is how little of the headline ingredient is present.

The arithmetic that a proprietary blend cannot escape

Here is the part almost no consumer guide bothers to show you. The descending-order requirement is not a formatting nicety. It is a mathematical constraint, and it lets you calculate a hard ceiling for every ingredient in a blend without any inside information at all.

If ingredients must be listed heaviest first, then the ingredient in position two can be no heavier than the ingredient in position one. The ingredient in position three can be no heavier than position two, and so on. It follows that the ingredient in position n can never exceed the blend total divided by n. If it did, the ingredients above it would each have to be larger still, and their combined weight would exceed the blend.

Applied to a hypothetical 500 mg blend containing six named ingredients, the ceilings look like this.

Pixie-dust arithmetic: a hypothetical 500 mg blend of six ingredients

Note what this table does and does not tell you. It does not prove that any particular blend is underdosed. It proves something narrower and more useful: that for anything listed below second or third position in a modest blend, an adequate dose is arithmetically impossible. The ceiling is fixed by the label itself. If a 500 mg blend lists rhodiola sixth and a brand implies you are getting the benefit seen in the fatigue trials, the label has already contradicted the claim.

The same reasoning cuts the other way, which is worth saying plainly. A large blend — say 2,000 mg — can carry genuinely adequate amounts of several ingredients. Blend size matters. "Proprietary blend" on its own is not a verdict.

The harder problem: disclosed doses that are still below trial doses

Proprietary blends get the attention because they are easy to be angry about. The more common and more consequential failure mode is quieter. A label discloses every ingredient and every milligram, in full public view, and the amounts are still well under what the research used.

This is not usually deception. Multi-ingredient formulas face a physical constraint: a capsule holds a finite amount of powder, and a two-capsule serving cannot contain trial doses of five different actives without becoming a scoop of powder. Something has to give. The honest version of that trade-off is stating it. The dishonest version is citing a trial at 300 mg while supplying 100 mg.

So here is the comparison that actually decides whether a label is meaningful — what the human trials used, with the study attached.

Disclosed dose vs the dose used in human trials

Two things fall out of that table immediately. First, the trial doses are not interchangeable with each other — a 100 mg dose is generous for L-theanine and negligible for rhodiola SHR-5. A label amount only means something next to the dose used for that ingredient. Second, several of the doses come with a named, characterised extract attached. SHR-5 and G115 are not generic material. Matching the milligram figure while using a different extract does not reproduce the trial, and a brand that cites Olsson 2009 while supplying an unspecified rhodiola powder is citing a study it has not replicated.

The three label patterns you will actually encounter

In practice, nootropic and cognitive-support labels fall into three shapes, and telling them apart takes seconds once you know what you are looking at.

Pattern one: full disclosure. Every ingredient carries its own milligram figure. There is no pooled total. A worked example, applied to the product this site covers, is in the full Memopezil ingredient list, dose by dose. This is the format that permits real comparison, and it is the format you should be able to check against the trial-dose table above. It is also, increasingly, a selling point — which is why some brands now advertise transparency while still omitting standardisation percentages, quietly reintroducing the same ambiguity one level down.

Pattern two: the hybrid. One or two flagship ingredients get named doses, and everything else is swept into a blend beneath them. This is the most common arrangement and the most revealing one. The named ingredients are the ones the brand is confident about; the pooled ones are, at minimum, the ones it would rather you did not price out. The hybrid label is not automatically dishonest, but the choice of what to disclose is itself information.

Pattern three: the total blend. A fanciful name, a single weight, and a list of ingredients underneath. Here the only tool you have is the descending-order ceiling. Apply it to the ingredient the marketing leans on hardest. If that ingredient sits fourth or fifth in a blend of a few hundred milligrams, the marketing and the label are describing different products.

One naming note worth carrying with you: "Proprietary Blend" is not a required phrase. The regulation permits "other appropriately descriptive term or fanciful name", which is why you will see Cognitive Matrix, Focus Complex, NeuroFormula and similar. A branded-sounding name with one milligram figure beside it and an indented ingredient list beneath it is a proprietary blend, whatever it is called.

Standardisation: the second number a label can quietly omit

Even a fully named dose can be close to meaningless without one more figure. "200 mg of bacopa monnieri extract" describes a weight of powder, not a quantity of active compound. Botanical extracts vary enormously in concentration depending on the plant part used, the solvent, and the extraction ratio. Two 200 mg extracts can differ several-fold in what actually matters.

Standardisation is the fix. A standardised extract declares the percentage of a marker compound — bacosides for bacopa, salidroside and rosavins for rhodiola, ginsenosides for ginseng, withanolides for ashwagandha. Mushroom labels have their own version of this problem, where the number that matters is the plant part and the beta-glucan percentage rather than a marker compound — the case we work through in the lion's mane fruiting body versus mycelium label guide. That percentage lets you convert a powder weight into an active-compound weight and compare products honestly.

It also does a second job that gets overlooked: for some plants, the choice of marker tells you which species you have. Rosavins are essentially specific to Rhodiola rosea, which is why a rhodiola label declaring salidroside alone leaves a real question unanswered. We unpack that specific case in detail in our guide to rosavins versus salidroside on a rhodiola label.

Three things to look for, in order of usefulness: the marker percentage, the plant part, and the extract ratio (a "10:1" extract means ten parts raw herb condensed into one part extract). A label with all three is being straight with you. A label with none of them is asking you to take the powder weight on faith.

A 7-point label-reading checklist

Run any Supplement Facts panel through these seven checks. It takes about ninety seconds and it settles most purchasing questions before price ever enters the conversation.

1. Blend or named doses? Look for the words "Proprietary Blend", "Matrix", "Complex" or a fanciful name with a single milligram figure beside it. If every ingredient has its own number, skip to point three.
2. Where does each ingredient sit in the order? Blend contents are listed heaviest first. Divide the blend total by the position number to get the ceiling for anything below the top slot.
3. Is the standardisation percentage declared? No marker percentage means the milligram figure cannot be converted into active content.
4. Is the extract ratio or plant part given? Root, leaf, whole herb and aerial parts are not equivalent, and neither are a 4:1 and a 20:1 extract.
5. Serving size versus capsules per bottle. A "60 capsule" bottle at three capsules per serving is a twenty-day supply. Divide before you compare prices.
6. How much of the capsule is "other ingredients"? Rice flour, magnesium stearate and silicon dioxide are normal in small amounts. A long list of bulking agents next to small active doses is informative.
7. Is there third-party testing? A named testing programme or a certificate of analysis available on request is a meaningfully higher bar than a facility claim alone.

What this test does not prove

It would be easy to finish here with a tidy conclusion that transparent labels are good and blends are bad. That would be overstating the evidence, so we will not.

A fully disclosed label proves exactly one thing: that you can check the numbers. It does not make the numbers adequate — an ingredient can be fully disclosed and still, as with the 540 mg BCAA blend we examine separately, have almost no human cognitive evidence behind it at any dose. It also says nothing about raw material quality, correct plant identification, contamination testing, or whether the capsules contain what the panel claims. Disclosure is a precondition for scrutiny, not a substitute for it.

The same test also has to be applied to products we cover, including this site's subject. Memopezil publishes named doses with no proprietary blend: bacopa monnieri extract 200 mg, rhodiola rosea extract 100 mg standardised to 3% salidroside, L-theanine 100 mg, panax ginseng extract 90 mg, and a 540 mg branched-chain amino acid blend. Run that through the table above and the picture is mixed and worth stating plainly. The L-theanine amount matches the 100 mg used by Kelly and colleagues. The bacopa amount is 200 mg against 300 mg in Stough 2001 and Calabrese 2008 — disclosed, checkable, and below the studied dose. The rhodiola figure is 100 mg standardised to salidroside, which is a different specification from the SHR-5 extract dosed at 576 mg/day in Olsson 2009 and at 170 mg/day in Darbinyan 2000. The ginseng amount is 90 mg against 200 mg of G115 in Reay 2005.

That is what an honest label read looks like when the answer is not flattering. A transparent panel earns the right to be checked; it does not pass the check automatically. And no multi-ingredient supplement, this one included, can transfer a single-ingredient trial result to the finished product — the finished formula has not itself been trialled, and any brand implying otherwise is over-reaching. For the underlying evidence on the two lead botanicals, see our review of what the bacopa and rhodiola research actually supports, and for the timeline question specifically, how long bacopa monnieri takes to work.

Frequently asked questions