What is actually in the vials sold as retatrutide?

Several independent laboratories have now tested products sold as retatrutide or semaglutide. The pattern is consistent: the amount is wrong more often than it is right, in both directions; the peptide is a small fraction of what is in the vial; and things that should never be in an injection have been found in it.

Who has tested these products

ABC News and the University of Queensland, June 2026. A vial sold online as 10 mg of retatrutide was tested by the University of Queensland's School of Public Health with an accredited Melbourne laboratory. It contained around 19 mg, nearly double the label, at 16.6% purity; the laboratory said the remainder was likely to be fill material such as mannitol. Gastroenterologist Dr Darcy Holt, who had treated a patient losing an estimated eight litres of fluid a day to diarrhoea, said "the potential for overdose is enormous".^[1]

ABC News and a Melbourne analytical laboratory, August 2026. The laboratory, Leeder Analytical, shared its 2026 results for 18 samples sold as retatrutide. One matched its label. About 30% were stronger than labelled, about 30% weaker, and one contained none of the peptide. Its chemist said the peptide usually made up about 10% of a vial's contents, leaving "another 90 per cent that's unaccounted-for, which could contain a different toxin or something else unsafe".^[2] A peer-reviewed analysis of products sold as retatrutide in Australia was published in Drug and Alcohol Review in 2026.^[5]

Semaglutide vials sold online, 2024. Researchers test-purchased semaglutide from online sellers without a prescription and analysed it. Purity was 7.7%, 9.0% and 14.4% against a claimed 99%. Every vial contained endotoxin, at up to 8.95 endotoxin units per milligram. Each contained 28.6% to 38.7% more semaglutide than its label. All were classed as substandard and falsified.^[3]

UK Yellow Card reports, 2023 onward. The MHRA confirmed that some fake Ozempic and Saxenda pens reported through the Yellow Card scheme contained insulin rather than the labelled drug; some people were hospitalised with hypoglycaemic shock.^[4] The World Health Organization issued an alert on falsified semaglutide in June 2024 covering batches found in the UK, Brazil and the United States.^[6]

MHRA raids, 2025 to 2026. Officers found tens of thousands of empty pens ready to be filled next to raw chemical ingredients.^[7] That is what hand-filling looks like, and it explains most of what follows. The enforcement page has the details.

The ways a vial can be wrong

Each of these is a separate failure. A vial can pass one and fail the rest.

Wrong identity

The vial contains something other than what the label says. The clearest UK example is insulin in pens sold as weight-loss injections.^[4] The Melbourne laboratory's sample with no retatrutide in it is the same failure in a different form.^[2] For the person: you get none of the intended effect and all of the effect of whatever is really there. Insulin in someone who does not need it can cause dangerously low blood sugar within an hour.

Wrong strength, in both directions

Too much is the finding that surprised the chemists: nearly double the label in the Queensland test, and roughly as many samples over the label as under it in the Melbourne series.^[1][2] For the person: retatrutide's side effects rise steeply with dose in the trials, where the dose was known and stepped up slowly under supervision. Twice the intended amount of a drug you have never taken before is the scenario behind the Australian cases of relentless diarrhoea, severe dehydration and acute kidney injury.^[1] Too little is not safe either: it means the next vial, from the same or a different batch, may be a large jump you did not plan for.

Low purity and the unknown remainder

Purity of 16.6% or 7.7% means that most of what is in the vial is not the peptide. Some of that is expected fill material such as mannitol. Nobody has accounted for the rest.^[1][3] For the person: you are injecting a mixture that is mostly unidentified. The Victorian liver-injury cluster is thought to involve a contaminant in products labelled retatrutide, which is what "unaccounted-for" can mean in practice.^[8] See the liver-injury cluster page.

Endotoxin

Endotoxin is fragments of the outer wall of certain bacteria. It survives filtering and most sterilisation, so a vial can be free of living bacteria and still contain it. Licensed injectables must pass a strict endotoxin limit. The 2024 semaglutide study found it in every sample.^[3] For the person: injected endotoxin causes fever, chills, inflammation and, at higher levels, a reaction that looks like septic shock.

Not sterile

Sterility cannot be judged by looking: a clear liquid can still carry bacteria or fungi. A product made up from powder in a warehouse and loaded into pens by hand has been through none of the validated sterile-filling steps a licensed medicine goes through.^[7] For the person: abscesses at the injection site, cellulitis spreading from it, and in the worst case sepsis, which is a 999 emergency.^[9] The injectables and infection page covers this in more detail.

Fragments and incomplete chains

Retatrutide is a peptide, a chain of amino acids built one link at a time. At each step a small fraction of chains fail to gain the next link or gain a damaged one, so the finished product always contains shortened or altered chains unless they are removed by careful purification at the end. Those near-identical chains can be hard to separate from the full peptide with standard HPLC methods, so a purity figure can count some of them as product. The authors of the 2024 semaglutide study noted that impurities from automated peptide synthesis are highly variable, that some are difficult to detect by mass spectrometry, and that the low purity they found suggested purification steps had been skipped or cut short because they are slow and expensive.^[3] For the person: fragments may be inactive, may act differently from the full peptide, or may provoke an immune reaction. Nobody has studied what they do in humans, because no licensed product contains them.

Why this is not a quality-control problem you can solve

The MHRA's summary is blunt: illegally sold products "may be fake, contaminated, incorrectly dosed, or contain powerful ingredients not listed on the packaging".^[10] Most seized medicines "can contain too much or too little of the declared active ingredient and may also contain other harmful ingredients".^[11]

The failures above are independent of each other. A test for one says nothing about the others, and a test on one vial says nothing about the next. That is why a certificate of analysis, which is usually a single HPLC purity figure on a single sample, does not answer the question. The certificate of analysis page goes through that in detail.