What Is Actually in the Vial

A vial of white powder sits on a bench. On the side is a label: a name, sometimes a sequence abbreviation, a figure in milligrams, often the word purity followed by a number with two decimal places. Yes — people have tested what is actually in vials like these. Official medicines control laboratories, doping-control scientists and forensic chemists have taken peptide preparations seized at borders or bought deliberately from illegal online sellers, run them through chromatography and mass spectrometry, and published what came out. The literature is small, scattered across analytical journals, and almost nobody outside those fields reads it. What it shows is that the label is a claim — not a lie, necessarily, and not a fact either, but a claim printed by whoever filled the vial, which in most of this market no independent party has confirmed.

This site has argued elsewhere that the claims made for research peptides outrun the evidence behind them — that a run of animal papers cannot carry a sentence about people. The question here is narrower, and worth keeping separate. Set aside whether any given peptide does anything at all. Ask only whether the material in the vial is the material named on it, in the amount stated, with nothing else of consequence in there. That is not settled by argument. Instruments answer it, and a few laboratories have pointed instruments at it.

A question nobody is obliged to answer

A licensed medicine carries a paper trail that most people never think about and which is, in engineering terms, the actual product. The identity of the active ingredient was confirmed against a reference standard, the content assayed, the impurities characterised and held below limits argued out in advance — and every one of those numbers belongs to the specific batch in the specific box. Material supplied for laboratory research sits outside that apparatus by definition. That is not an accusation; it is what the category means. A research reagent is sold on the understanding that a competent laboratory will characterise it as part of doing the work. Nobody is obliged to verify anything on the buyer's behalf.

Which is why the analytical literature looks the way it does. Nobody funded a survey. The papers come from institutions that had another reason to open the boxes: customs officers with a seizure, anti-doping laboratories working out what athletes were being sold, inspectorates building screening methods for suspicious parcels. One Belgian control laboratory published exactly that — a chromatography–mass spectrometry screen detecting twenty-five peptides in half an hour, with ten of them quantified, built around substances European authorities had already confiscated 1. It is a tool rather than a survey, but that a national laboratory needed the tool is itself informative.

What the analysts actually found

The most detailed look at peptide products came from the same Belgian group a few years later. They took the ten most frequently encountered falsified peptide drugs on their market, acquired from three suspected illegal internet pharmacies, and profiled them fully: active ingredient, peptide-related impurities, small-molecule contaminants, elemental impurities, residual solvents. Purity of the cysteine-containing peptides ranged between 5% and 75%, and the amount of drug per unit varied widely. Several samples carried arsenic at up to ten times the relevant toxicity limit, all of it in the more toxic inorganic form; one carried lead. The authors read the impurity pattern as the signature of purification steps omitted or only partly carried out 2.

A study of a tanning peptide sold as an injectable online isolates the quantity problem. Researchers bought vials from three shops, each stating 10 mg per vial, and measured between 4.32 and 8.84 mg. Products from two shops contained unidentified impurities at 4.1% to 5.9%; in the third, impurities sat below the limit of quantification 3. That third result deserves as much attention as the others. The honest summary is variability, not universal adulteration.

Then the category where the molecule is not the one named. Analysing new growth-promoting black-market products, one group characterised a growth hormone variant of 192 amino acids — the natural protein has 191 — carrying an extra alanine at one end, together with three peptide analogues each consisting of a known compound extended by a single glycine residue 4. To an instrument these are unambiguous: different mass, different fragmentation. To a buyer they are invisible, arriving in a vial bearing the name of the compound they are not.

For scale, though a blunt kind, a German doping-control laboratory reported on 337 black-market products examined between 2010 and 2013, identifying 67 active ingredients, 49 of them classed as doping agents by the World Anti-Doping Agency. Peptide hormones and growth factors made up 12.8%; the bulk was steroidal 5. It is not a peptide study and should not be read as one.

The most recent entry concerns a peptide that is a licensed medicine, bought from sellers operating outside prescription. Investigators mapped the online market, made test purchases from illegal pharmacies, and received three injection vials; the prefilled pens ordered never arrived at all. Content in all three exceeded the labelled amount, by 28.56% to 38.69%. Measured purity ran from 7.7% to 14.37%, against a claimed 99%. Every sample tested positive for endotoxin, at 2.1645 to 8.9511 endotoxin units per milligram, while bacterial culture grew nothing 6. Three vials is three vials, and the study says so. But note the direction: these contained more than promised, not less. Both directions are the same failure.

The several different ways a vial can be wrong

It helps to separate these, because they have different causes and different tests would catch them. Lumped together as fake, they become one scare story. Taken apart, they are ordinary manufacturing failures, each boring and specific.

That last row carries a point the semaglutide analysis makes vividly. Those samples grew no viable organisms and still tested positive for endotoxin in every case 6. Endotoxin is a fragment of bacterial cell wall, and it survives the death of the bacterium that produced it — sterilisation kills the organism and leaves the fragment behind. A sterility test and an endotoxin test answer different questions, and passing one tells you nothing about the other.

Why a peptide preparation is never one molecule

The impurity row deserves unpacking, because it is not a defect of bad manufacturers. It is a property of the chemistry. Peptides are built by solid-phase synthesis: the chain is anchored to a bead of resin and extended one amino acid at a time, each residue coupled to the growing end, washed, deprotected, and the cycle repeated. Every coupling step is efficient. None is perfect. A small fraction of chains fails to receive its residue on any given cycle, then carries on growing permanently one amino acid short. That is a deletion sequence — the characteristic impurity of the method.

Follow the arithmetic and the implication is unavoidable. Even at very high per-step efficiency the losses compound across every cycle, so the proportion of chains that are complete falls steadily as the peptide gets longer. What comes off the resin is never a single molecule. It is a distribution: mostly the intended sequence, alongside a population of near-misses each missing something, plus truncated fragments. This is true of every peptide made this way, including those inside licensed medicines.

What separates a pharmaceutical-grade peptide from a poor one is not that the first avoided the problem, but what happened next. Purification — chromatographic separation of the intended chain from its near-relatives — narrows the distribution, and it is slow, expensive, wasteful of material, and entirely invisible in the finished powder. It is precisely the step a manufacturer under cost pressure can quietly shorten, which is the inference the Belgian analysts drew from the impurity fingerprints in front of them 2. Two vials of identical-looking white powder can have had wholly different amounts of care applied to them, and nothing in their appearance will say which is which.

How to read a certificate of analysis

This is where the certificate of analysis enters, and it deserves fair handling. A certificate is a real analytical document and there is nothing inherently suspect about one. But certificates differ enormously in what they establish, and the difference is legible on the page. The question is not whether a certificate exists. It is what the document, read as a document, is in a position to tell you.

• Batch-specific? A certificate naming the batch it describes, matching the batch on the vial in front of you, is making a claim about that material. One with no batch identifier is describing something else.
• Dated? An undated certificate cannot be placed in time or connected to a production run.
• Methods named? Meaningful documents say how each figure was obtained. A bare number with no method has no provenance.
• Identity separated from purity? A purity percentage says how much of the sample is one substance. It does not say that substance is the one on the label.
• Attributable? A named laboratory, a signature, a contact — an analysis somebody will stand behind — is stronger than an anonymous page.
• Independent? Third-party testing and a seller's own untested figures are different kinds of evidence. That is a point about who checks whom, not about anyone's honesty.

None of this makes a strong certificate a guarantee. Even the best describes a defined quantity of material at a defined moment under stated methods, and says nothing about that vial's journey afterwards — the warehouse, the courier, the summer, the customs shelf. That is a great deal more than nothing and considerably less than certainty, and knowing which of the two you are holding is the whole skill.

Why this sits upstream of every other argument

Here is the point that gives this piece its reason to exist. Every discussion about whether a research peptide works — every study, every argument, every anecdote — rests on an assumption almost never stated aloud: that the material involved was the material described. Remove it and the conversation loses its footing. A result showing no effect might mean the compound does nothing. It might equally mean the vial did not contain the compound, or held a fraction of the stated amount, or held a molecule one residue longer. A result showing an effect is no safer: the effect might belong to something never named on the label.

So identity and purity are not a side issue about consumer protection, to be dealt with once the interesting scientific questions are settled. They are logically prior to them. An experiment on an uncharacterised substance does not produce a weak answer about that substance; it produces no answer, because there is no defined thing the answer would be about. Uncertainty at this layer propagates upward into everything built on it, and no amount of care downstream repairs it.

What this literature can and cannot tell you

The temptation now is to reach for a proportion — to say that some share of what circulates is wrong, and to feel informed. The evidence does not support that move, and it is worth saying why rather than simply declining.

These studies did not sample randomly. They analysed what customs had seized, what anti-doping laboratories were handed, what investigators bought on purpose from sellers already flagged as illegitimate 156. Every one of those routes selects for the suspicious — right for the purposes the work was done for, wrong for estimating a base rate. Sample sizes are small: ten products in one study, three vials in another, three shops in a third 236. The papers are separated by a decade, several jurisdictions and different classes of compound. Aggregating them into one figure would produce a statistic with the appearance of authority and nothing behind it.

The defensible claim is modest, and worth making precisely. This is not hypothetical and not rumour: preparations sold outside regulated supply have been analysed by competent laboratories, in peer-reviewed work, and found to contain the wrong quantity, substantial impurity, toxic elemental contamination, bacterial endotoxin, and in some cases a different molecule from the one named 2346. Each finding is documented. None tells you how often it happens, and anyone offering that figure has gone beyond what has been established.

That is neither alarming nor reassuring, which is probably the right temperature. What the analytical literature provides is not a risk estimate but a change of question. The thing to carry away is not a percentage but a habit: noticing that a label is a claim rather than a fact, that claims are the sort of thing that can be checked, and that whether anyone has checked this one has an answer — sometimes yes, often no, almost always findable. The gap between a label and a molecule is rarely dramatic. It is just unexamined, which is a quieter problem and a more common one.