Skip to content
Calculator

Verify

Independent lab results

Published results from laboratories that tested peptide products, most obtained by a study team itself and some submitted by buyers; each row's source says which, and a submitted vial's origin is only what the buyer declared, failures included. Every row names the laboratory, the method, when the testing was done, what was expected, what was found, and where it was published, so you can check it.

A seller's own certificate is not a source here, and neither is a result nobody can trace. The table shows no prices, no purchase links and no overall score for any seller: a row is one measurement of one product, not a verdict on a company. Data as of November 2, 2026; 40 results. Rules: methodology. Download: lab-results.json.

The big picture: an April 2026 preprint analysed 6,285 test results on 14 research peptides from 203 companies; 41.6% to 71.1% failed, mostly on the amount in the vial. It reports aggregates, not products, so it is summarised on What lab testing shows rather than added here.

By compound

How many products of each compound published tests cover, and what they found. A product is one vial, pen or pack as sold; a product tested for identity and for content counts once. Results are counted per product, so one failed content test on three vials counts three. A product counts as failed when a test found the wrong molecule, less or more than the label, or contamination; "inconclusive" means the source gave no standard to judge the figure against.

CompoundProducts testedPublished reportsResults by measure (products)Failed results
ACE-031141identity: 14 fail14
BPC-15711identity: 1 pass0
Epithalon21identity: 2 pass0
Follistatin-344171identity: 8 fail, 9 pass8
GHRP-221identity: 1 inconclusive, 1 pass0
Hexarelin11identity: 1 pass0
IGF-1 LR311identity: 1 fail1
Melanotan II42identity: 4 pass; content: 4 fail; purity: 3 inconclusive4
MK-67741identity: 4 pass0
Retatrutide31identity: 3 pass; content: 3 fail3
Semaglutide73identity: 2 fail; content: 3 fail, 3 inconclusive, 2 pass; purity: 2 inconclusive; sterility: 3 pass; endotoxin: 3 inconclusive5
TB-50011identity: 1 pass0

What a forged or misleading certificate looks like, with sourced examples: Fake COA red flags. How peptides are made and purified, and which impurities each step can leave: how peptide drugs are made.

Results

40 of 40 shown
CompoundSellerMeasureResultOutcomeMethodLaboratoryTestedSource
SemaglutideBiotech Peptidescontent3.86 mg
expected: 3 mg of semaglutide per vial, as labelled
failLC-MSMass Spectrometry Core Facility, Szentágothai Research Centre, University of PécsAugust 2023 to March 20241PMID 39509151; PMID 39093567
SemaglutideBiotech Peptidescontent7.7 %
expected: no share of the powder is stated; the seller claimed at least 99% purity, which is a different measure
inconclusiveLC-MSMass Spectrometry Core Facility, Szentágothai Research Centre, University of PécsAugust 2023 to March 20242PMID 39509151; PMID 39093567
SemaglutideBiotech Peptidessterilityno viable microorganisms
expected: sterile, per Ph. Eur. 2.6.1 and USP <71>
passUSP-71PharmaValid Ltd microbiology laboratory, Budapest (ISO 14644-1 certified)August 2023 to March 20243PMID 39509151; PMID 39093567
SemaglutideBiotech Peptidesendotoxin<2.1645 EU/mg
expected: no acceptance limit is stated in the source for this product
inconclusiveLALPharmaValid Ltd microbiology laboratory, Budapest (ISO 14644-1 certified)August 2023 to March 20244PMID 39509151; PMID 39093567
SemaglutideSemaSpacecontent2.77 mg
expected: 2 mg of semaglutide per vial, as labelled
failLC-MSMass Spectrometry Core Facility, Szentágothai Research Centre, University of PécsAugust 2023 to March 20245PMID 39509151; PMID 39093567
SemaglutideSemaSpacecontent14.37 %
expected: no share of the powder is stated; the seller claimed at least 99% purity, which is a different measure
inconclusiveLC-MSMass Spectrometry Core Facility, Szentágothai Research Centre, University of PécsAugust 2023 to March 20246PMID 39509151; PMID 39093567
SemaglutideSemaSpacesterilityno viable microorganisms
expected: sterile, per Ph. Eur. 2.6.1 and USP <71>
passUSP-71PharmaValid Ltd microbiology laboratory, Budapest (ISO 14644-1 certified)August 2023 to March 20247PMID 39509151; PMID 39093567
SemaglutideSemaSpaceendotoxin8.9511 EU/mg
expected: no acceptance limit is stated in the source for this product
inconclusiveLALPharmaValid Ltd microbiology laboratory, Budapest (ISO 14644-1 certified)August 2023 to March 20248PMID 39509151; PMID 39093567
SemaglutideUSChemLabscontent1.34 mg
expected: 1 mg of semaglutide per vial, as labelled
failLC-MSMass Spectrometry Core Facility, Szentágothai Research Centre, University of PécsAugust 2023 to March 20249PMID 39509151; PMID 39093567
SemaglutideUSChemLabscontent8.97 %
expected: no share of the powder is stated; the seller claimed at least 99% purity, which is a different measure
inconclusiveLC-MSMass Spectrometry Core Facility, Szentágothai Research Centre, University of PécsAugust 2023 to March 202410PMID 39509151; PMID 39093567
SemaglutideUSChemLabssterilityno viable microorganisms
expected: sterile, per Ph. Eur. 2.6.1 and USP <71>
passUSP-71PharmaValid Ltd microbiology laboratory, Budapest (ISO 14644-1 certified)August 2023 to March 202411PMID 39509151; PMID 39093567
SemaglutideUSChemLabsendotoxin<2.8658 EU/mg
expected: no acceptance limit is stated in the source for this product
inconclusiveLALPharmaValid Ltd microbiology laboratory, Budapest (ISO 14644-1 certified)August 2023 to March 202412PMID 39509151; PMID 39093567
RetatrutidePARTY-04identityretatrutide confirmed in all three
expected: retatrutide, as labelled (expected molecular weight 4730.477 Da)
passMSAn independent NATA-accredited laboratory, not named in the paper (samples via PEDTest Australia)202613PMID 42559975
RetatrutidePARTY-05content5.13 mg
expected: 10 mg of retatrutide per vial, as labelled
failMSAn independent NATA-accredited laboratory, not named in the paper (samples via PEDTest Australia)202614PMID 42559975
RetatrutidePARTY-06content19.0 mg
expected: 10 mg of retatrutide per vial, as labelled
failMSAn independent NATA-accredited laboratory, not named in the paper (samples via PEDTest Australia)202615PMID 42559975
RetatrutidePARTY-07content16.5 mg
expected: 10 mg of retatrutide per vial, as labelled
failMSAn independent NATA-accredited laboratory, not named in the paper (samples via PEDTest Australia)202616PMID 42559975
Melanotan IIPARTY-08identitymelanotan II confirmed
expected: melanotan II, as sold
passLC-MSDepartment of Clinical Biochemistry, Vendsyssel Hospital (Aalborg University), Denmark201417PMID 24771717
Melanotan IIPARTY-08content4.32 to 8.84 mg
expected: 10 mg of melanotan II per vial, as each shop claimed
failHPLCDepartment of Clinical Biochemistry, Vendsyssel Hospital (Aalborg University), Denmark201418PMID 24771717
Melanotan IIPARTY-08purityunknown impurities 4.1 to 5.9% (two shops); below the quantification limit (one shop) %
expected: no stated purity; no pharmacopoeial standard exists for melanotan II
inconclusiveHPLCDepartment of Clinical Biochemistry, Vendsyssel Hospital (Aalborg University), Denmark201419PMID 24771717
SemaglutidePARTY-09content101 %
expected: 1 mg/mL of semaglutide, as labelled (100%)
passHPLCNovo Nordisk A/S analytical laboratories (CMC Analysis & Characterisation), Denmark202520PMID 42533250
SemaglutidePARTY-09puritysum of impurities 1.3% %
expected: no limit stated for a compounded product; the originator measured 3.2 to 3.4% in the same study
inconclusiveHPLCNovo Nordisk A/S analytical laboratories (CMC Analysis & Characterisation), Denmark202521PMID 42533250
SemaglutidePARTY-10content102 %
expected: 1 mg/mL of semaglutide, as labelled (100%)
passHPLCNovo Nordisk A/S analytical laboratories (CMC Analysis & Characterisation), Denmark202522PMID 42533250
SemaglutidePARTY-10puritysum of impurities 2.5% %
expected: no limit stated for a compounded product; the originator measured 3.2 to 3.4% in the same study
inconclusiveHPLCNovo Nordisk A/S analytical laboratories (CMC Analysis & Characterisation), Denmark202523PMID 42533250
Follistatin-344PARTY-11identityfollistatin 344 present in 9 of the 17 products
expected: follistatin 344, as each product was sold (the paper's E. coli FS344 standard runs at about 43.5 kDa)
passimmunoblot
identity only; does not measure purity or amount
Doping Control Laboratory Seibersdorf (Seibersdorf Labor GmbH), Austria201924PMID 31758732
Follistatin-344PARTY-12identityno follistatin in 8 of the 17 products: no protein larger than 20 kDa
expected: follistatin 344, as each product was sold (the paper's E. coli FS344 standard runs at about 43.5 kDa)
failSDS-PAGE
identity only; does not measure purity or amount
Doping Control Laboratory Seibersdorf (Seibersdorf Labor GmbH), Austria201925PMID 31758732
BPC-157PARTY-13identityBPC 157 confirmed
expected: BPC 157 (GEPPPGKPADDAGLV), as printed on the vial
passLC-MSSports Medicine Research and Testing Laboratory, Salt Lake City, Utah, USA201626PMID 28035768
TB-500PARTY-14identityAc-LKKTETQ (N-acetylated thymosin beta-4 17-23) only; no full-length thymosin beta-4
expected: TB-500 as this encyclopedia defines it: the acetylated LKKTETQ fragment of thymosin beta-4 (the product itself gave no sequence or molecular weight)
passLC-MSDoping Control Laboratory (DoCoLab), Ghent University, Belgium201227PMID 22962027
MK-677PARTY-15identityibutamoren detected in all 4 products whose label declared it
expected: ibutamoren (MK-677), as declared on each product's label
passLC-MSLGC, Fordham, Cambridgeshire, UK202028PMID 32748554
ACE-031PARTY-16identityno ACE-031 in 12 of the 14 products: full-length activin receptor IIB without the Fc fragment
expected: ACE-031 (ramatercept), as each product was sold: an activin receptor IIB fragment fused to the Fc part of human IgG1
failLC-MSDoping Control Laboratory Seibersdorf (Seibersdorf Labor GmbH), Austria202529PMID 40312924
ACE-031PARTY-17identityno ACE-031 in 1 of the 14 products: follistatin 344 instead
expected: ACE-031 (ramatercept), as each product was sold: an activin receptor IIB fragment fused to the Fc part of human IgG1
failimmunoblot
identity only; does not measure purity or amount
Doping Control Laboratory Seibersdorf (Seibersdorf Labor GmbH), Austria202530PMID 40312924
ACE-031PARTY-18identityno ACE-031 in 1 of the 14 products: no protein found
expected: ACE-031 (ramatercept), as each product was sold: an activin receptor IIB fragment fused to the Fc part of human IgG1
failSDS-PAGE
identity only; does not measure purity or amount
Doping Control Laboratory Seibersdorf (Seibersdorf Labor GmbH), Austria202531PMID 40312924
HexarelinPARTY-19identityhexarelin found, as labelled
expected: hexarelin, as printed on the vial
passLC-MSCologne Doping Control Laboratory (Institute of Biochemistry, German Sport University Cologne), Germany200932PMID 21204286
GHRP-2PARTY-20identityGHRP-2 found (about 50 µg per tablet)
expected: GHRP-2, as named on the label of an over-the-counter tablet supplement
passLC-MSCologne Doping Control Laboratory (Institute of Biochemistry, German Sport University Cologne), Germany200933PMID 21204286
GHRP-2PARTY-21identityGHRP-2 found (about 9 mg per ampoule)
expected: 'active GHRP-2 ester Z11', as named on the label of an over-the-counter drinking solution
inconclusiveLC-MSCologne Doping Control Laboratory (Institute of Biochemistry, German Sport University Cologne), Germany200934PMID 21204286
EpithalonPARTY-22identityEpitalon (AEDG) confirmed
expected: Epitalon, sequence AEDG, as printed on the vial ('50 mg Epitalon')
passLC-MSOfficial Medicines Control Laboratory, Scientific Institute of Public Health (WIV-ISP, now Sciensano), Brussels, Belgium201435PMID 25535022
EpithalonPARTY-23identityEpitalon (AEDG) confirmed
expected: Epitalon, as printed on the container ('EPITALON, food supplement', 3 mg of peptide per tablet)
passLC-MSOfficial Medicines Control Laboratory, Scientific Institute of Public Health (WIV-ISP, now Sciensano), Brussels, Belgium201436PMID 25535022
IGF-1 LR3PARTY-24identityno IGF-1 of any form; epidermal growth factor (EGF, 6215.9 Da) found instead
expected: long R3 IGF-1, as printed on the vial ('reconbinant IGF-1 long R3', misspelled on the label)
failLC-MSOfficial Medicines Control Laboratory, Scientific Institute of Public Health (WIV-ISP), Brussels, Belgium201437PMID 27390261
Melanotan IIPARTY-25identitymelanotan II confirmed
expected: melanotan II, as printed on the vial ('Melanotan II', 10 mg)
passLC-MSLaboratory of Clinical, Forensic and Environmental Toxicology, University Hospital of Liège, Belgium202438PMID 39302005
Melanotan IIPARTY-25content2.61 mg
expected: 10 mg of melanotan II per vial, as printed on the label
failHPLCLaboratory of Clinical, Forensic and Environmental Toxicology, University Hospital of Liège, Belgium202439PMID 39302005
SemaglutidePARTY-26identityno semaglutide; insulin glulisine found instead
expected: semaglutide, as the pens were presented (falsified Ozempic pre-filled pens)
failLC-MSCVUA Karlsruhe (Chemisches und Veterinäruntersuchungsamt Karlsruhe), GermanyNovember 202340CVUA Karlsruhe report

Two methods here show identity only: an immunoblot (western blot) shows whether a protein that a specific antibody recognises is present at the expected size, and an SDS-PAGE gel with a protein stain shows whether any protein of that size is there at all. Neither measures purity or how much is in the vial, which takes HPLC or mass spectrometry, so a row using either is an identity result and is marked so.

Notes on each result

  1. LR-0001: Measured 3.86 mg (SD 0.14) against 3 mg on the label: 28.56% more than labelled, which is the paper's own labelling-accuracy figure. The same method read the genuine Ozempic 1 mg pen at +5.05% over its label. A reader dosing by the label would receive more than intended. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  2. LR-0002: Semaglutide made up 7.7% (SD 0.28) of the powder in the vial by weight (3.86 mg of 50.1 mg). The paper's table heads this column 'measured purity' beside the 99% claimed, but its text calls the figure the polypeptide concentration, and the chromatogram showed only semaglutide, with no peptide-like impurities. So this is not the chromatographic purity a research-grade '99%' claim means: the rest of the powder is undeclared material the paper did not identify. The measurement stands; the comparison with 99% is not like-for-like. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  3. LR-0003: Tested by what the paper calls a direct injection technique, under both pharmacopoeias. The paper reports the lyophilised powder free of viable microorganisms at the time of testing; it does not test the vial after reconstitution or storage. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  4. LR-0004: Reported as below 2.1645 EU/mg by kinetic turbidimetric assay (Ph. Eur. 2.6.14, USP <85>) — an upper bound, not a measured level, although the paper's abstract describes endotoxin as detected in all samples. No limit is stated, so no pass or fail is given here. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  5. LR-0005: Measured 2.77 mg (SD 0.12) against 2 mg on the label: 38.69% more than labelled, which is the paper's own labelling-accuracy figure. The same method read the genuine Ozempic 1 mg pen at +5.05% over its label. A reader dosing by the label would receive more than intended. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  6. LR-0006: Semaglutide made up 14.37% (SD 0.63) of the powder in the vial by weight (2.77 mg of 19.3 mg). The paper's table heads this column 'measured purity' beside the 99% claimed, but its text calls the figure the polypeptide concentration, and the chromatogram showed only semaglutide, with no peptide-like impurities. So this is not the chromatographic purity a research-grade '99%' claim means: the rest of the powder is undeclared material the paper did not identify. The measurement stands; the comparison with 99% is not like-for-like. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  7. LR-0007: Tested by what the paper calls a direct injection technique, under both pharmacopoeias. The paper reports the lyophilised powder free of viable microorganisms at the time of testing; it does not test the vial after reconstitution or storage. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  8. LR-0008: Endotoxin measured at 8.9511 EU/mg by kinetic turbidimetric assay (Ph. Eur. 2.6.14, USP <85>), the highest of the three; the paper says it indicates potential environmental contamination during production. No limit is stated, so no pass or fail is given here. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  9. LR-0009: Measured 1.34 mg (SD 0.07) against 1 mg on the label: 33.58% more than labelled, which is the paper's own labelling-accuracy figure. The same method read the genuine Ozempic 1 mg pen at +5.05% over its label. A reader dosing by the label would receive more than intended. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  10. LR-0010: Semaglutide made up 8.97% (SD 0.51) of the powder in the vial by weight (1.34 mg of 14.9 mg). The paper's table heads this column 'measured purity' beside the 99% claimed, but its text calls the figure the polypeptide concentration, and the chromatogram showed only semaglutide, with no peptide-like impurities. So this is not the chromatographic purity a research-grade '99%' claim means: the rest of the powder is undeclared material the paper did not identify. The measurement stands; the comparison with 99% is not like-for-like. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  11. LR-0011: Tested by what the paper calls a direct injection technique, under both pharmacopoeias. The paper reports the lyophilised powder free of viable microorganisms at the time of testing; it does not test the vial after reconstitution or storage. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  12. LR-0012: Reported as below 2.8658 EU/mg by kinetic turbidimetric assay (Ph. Eur. 2.6.14, USP <85>) — an upper bound, not a measured level, although the paper's abstract describes endotoxin as detected in all samples. No limit is stated, so no pass or fail is given here. Named under the sample-control rule: the study team bought this product itself, and the paper reports its results separately from the other products. Test date: The research letter on this study states that test purchases and analytical testing were performed from August 2023 to March 2024; neither paper gives a date for any individual test.
  13. LR-0013: Identity was confirmed by molecular weight in all three products sold as retatrutide. The three products came from sellers the letter does not name; this row summarises all three, so PARTY-04 stands for that group, not for one seller. A buyer submitted each sample, so where each vial came from is self-declared. Test date: The letter does not date the analyses. It was received by the journal on July 15, 2026, so the testing was done on or before then; recorded as 2026, the year of the letter.
  14. LR-0014: Sample 1 of the letter's Table 1: 5.13 mg against 10 mg on the label (51.3% of the labelled content). The letter says only that peptide content was quantified to estimate the amount per vial; the method is recorded as mass spectrometry because identity was confirmed by molecular weight. A reader dosing by the label would receive about half the intended amount. A buyer submitted the sample, so its origin is self-declared. Test date: The letter does not date the analyses. It was received by the journal on July 15, 2026, so the testing was done on or before then; recorded as 2026, the year of the letter.
  15. LR-0015: Sample 2 of the letter's Table 1: 19.0 mg against 10 mg on the label (190.0% of the labelled content). The letter says only that peptide content was quantified to estimate the amount per vial; the method is recorded as mass spectrometry because identity was confirmed by molecular weight. A reader dosing by the label would receive far more than the intended amount. A buyer submitted the sample, so its origin is self-declared. Test date: The letter does not date the analyses. It was received by the journal on July 15, 2026, so the testing was done on or before then; recorded as 2026, the year of the letter.
  16. LR-0016: Sample 3 of the letter's Table 1: 16.5 mg against 10 mg on the label (165.0% of the labelled content). The letter says only that peptide content was quantified to estimate the amount per vial; the method is recorded as mass spectrometry because identity was confirmed by molecular weight. A reader dosing by the label would receive far more than the intended amount. A buyer submitted the sample, so its origin is self-declared. Test date: The letter does not date the analyses. It was received by the journal on July 15, 2026, so the testing was done on or before then; recorded as 2026, the year of the letter.
  17. LR-0017: Identity was verified by retention time and the abundance ratios of five fragment ions. The study bought vials from three online shops and reports by shop, not by vial; samples counts one per shop, and PARTY-08 stands for the three shops together, not for one seller. Test date: The abstract gives no testing date. The paper was first published online on April 25, 2014, so the analyses were done before then; recorded as 2014, the year of first publication.
  18. LR-0018: Quantified by LC-UV at 218 nm with a method validated against the guidelines for authorised medicines. Every shop claimed 10 mg per vial; the vials held 4.32 to 8.84 mg, so all fell short. Three shops, one per sample, reported as a range. Test date: The abstract gives no testing date. The paper was first published online on April 25, 2014, so the analyses were done before then; recorded as 2014, the year of first publication.
  19. LR-0019: Vials from two shops contained unidentified impurities of 4.1 to 5.9%; impurities in the third shop's vials were below the quantification limit. Read as inconclusive because no standard sets a limit for an unapproved product. Test date: The abstract gives no testing date. The paper was first published online on April 25, 2014, so the analyses were done before then; recorded as 2014, the year of first publication.
  20. LR-0020: Product CP12 in the paper: an injectable compounded semaglutide in a vial from a US outsourcing facility, labelled 1 mg/mL. Before light exposure it measured 1.01 mg/mL (101% of the label). After ICH Q1B light exposure it fell to 0.95 mg/mL (95%), while the originator did not change. The study was funded by, and run in the laboratories of, Novo Nordisk, which makes the originator; read it with that in mind. Test date: The paper does not date the tests. It says the products 'were commercially available for patient use at the time of testing', and its references were accessed on December 11, 2025; recorded as 2025.
  21. LR-0021: Product CP12: the sum of impurities was 1.3% before light exposure and 5.1% after it; high-molecular-weight proteins rose from <0.3% to 2.6%. As received it was within the originator's own range, so the reading is inconclusive rather than a failure; the change under light is the paper's finding. Manufacturer-funded study, as noted on the content row. Test date: The paper does not date the tests. It says the products 'were commercially available for patient use at the time of testing', and its references were accessed on December 11, 2025; recorded as 2025.
  22. LR-0022: Product CP13 in the paper: an injectable compounded semaglutide in a vial from a US outsourcing facility, labelled 1 mg/mL. Before light exposure it measured 1.02 mg/mL (102% of the label). After ICH Q1B light exposure it fell to 0.86 mg/mL (86%), while the originator did not change. The study was funded by, and run in the laboratories of, Novo Nordisk, which makes the originator; read it with that in mind. Test date: The paper does not date the tests. It says the products 'were commercially available for patient use at the time of testing', and its references were accessed on December 11, 2025; recorded as 2025.
  23. LR-0023: Product CP13: the sum of impurities was 2.5% before light exposure and 14.2% after it; high-molecular-weight proteins rose from 0.3% to 14.1%. As received it was within the originator's own range, so the reading is inconclusive rather than a failure; the change under light is the paper's finding. Manufacturer-funded study, as noted on the content row. Test date: The paper does not date the tests. It says the products 'were commercially available for patient use at the time of testing', and its references were accessed on December 11, 2025; recorded as 2025.
  24. LR-0024: Nine of the 17 products sold as follistatin 344 showed a main band at the FS344 position on SDS-PAGE, and immunoblotting with an anti-follistatin antibody confirmed it was follistatin. All nine also reacted with an anti-His-tag antibody, so they carried a His-tag, a laboratory purification tag the human protein does not have, and all nine held a high share of oligomers (mostly dimers and trimers) 'aside from impurities'. Identity only: an immunoblot shows that an antibody-reactive protein of the expected size is present; it does not measure purity or how much is in the vial, and the paper reports neither. The products came from suppliers in Bulgaria, Canada, China, India, Slovakia, Switzerland, the Netherlands, the UK and the USA, which the paper keeps anonymous (BM01 to BM17); PARTY-11 stands for these nine together, not for one seller. Test date: The paper does not date the analyses. It was published in the November 2019 issue of Drug Testing and Analysis, so the products were tested on or before then; recorded as 2019, the year of publication.
  25. LR-0025: Eight of the 17 products sold as follistatin 344 contained no protein larger than 20 kDa on SDS-PAGE with Coomassie stain (10 micrograms loaded per lane), so no follistatin. Mass spectrometry found small growth-promoting peptides in some of these 'empty' products, the paper's examples being MGF and GHRP-2; it does not show those data or say which products. Identity only: the gel shows that no protein of follistatin's size was there; it does not identify or measure what the vials held instead. PARTY-12 stands for these eight anonymous suppliers together, not for one seller. Test date: The paper does not date the analyses. It was published in the November 2019 issue of Drug Testing and Analysis, so the products were tested on or before then; recorded as 2019, the year of publication.
  26. LR-0026: The vial was labelled 'BPC 157, GEPPPGKPADDAGLV'. High-resolution accurate-mass LC-MS/MS found the doubly charged parent ion at m/z 710.3575 against 710.3594 expected (2.7 ppm) and eight fragment ions matching the sequence, each within 10 ppm. The paper reports identity only: it gives no purity and no amount for the vial. Where the vial was bought is unknown. Test date: The paper does not date the analysis. It was first published online on December 29, 2016, so the vial was analysed on or before then; recorded as 2016.
  27. LR-0027: The sealed vial held about 10 mg of white powder. HPLC with high-resolution (Orbitrap) mass spectrometry showed one main peak of molecular weight 888.49, the N-terminally acetylated LKKTETQ heptapeptide, confirmed by de novo sequencing, NMR and comparison with a synthesised reference; it did not contain full-length thymosin beta-4 (molecular weight 4,963). A slower gradient still showed a single peak, and no other active substance or excipient was detected, but the paper gives no purity figure or amount, so this is an identity result only. A buyer expecting the whole thymosin beta-4 protein would not have received it. Test date: The paper does not date the analysis. It was published online on September 7, 2012, so the vial was analysed on or before then; recorded as 2012.
  28. LR-0028: Four of the 20 capsule products declared ibutamoren (products 1, 4, 8 and 17, from four different manufacturers); UHPLC with high-resolution accurate-mass MS, against a reference standard, found it in all four. Three of the four were multi-ingredient 'stacks', and two of those held a discrepancy elsewhere: product 8 lacked its declared RAD140 and held undeclared YK11, and product 4 held undeclared RAD140. Ibutamoren was also found, undeclared, in four other products (2, 6, 7 and 14). The paper plots measured amounts against label claims only as a figure, so no amount is recorded here; identity only. PARTY-15 stands for the four manufacturers together, not for one seller. Test date: The paper does not date the analyses. It was first published online in August 2020, so the products were tested on or before then; recorded as 2020.
  29. LR-0029: Twelve of the 14 products sold as ACE-031 showed a main band at about 58.4 kDa on SDS-PAGE with many other proteins beside it, and an anti-ACVR2B immunoblot confirmed activin receptor IIB was present; neither test can tell ACE-031 from the receptor itself. In-gel tryptic digestion with LC-high-resolution MS found peptides from the receptor's cytoplasmic region and none from the human IgG1 Fc fragment ACE-031 carries, and IdeS protease, which cuts an Fc fusion protein, did not cleave the products, so the paper concludes the 12 held the full-length receptor, not ACE-031. Eleven of the 12 also carried a His-tag (all except BM03). Identity only: no purity or amount was measured. PARTY-16 stands for these 12 anonymous suppliers together, not for one seller. Test date: The paper does not date the analyses. It was published in 2025 (Drug Testing and Analysis volume 17, issue 10; PubMed date May 1, 2025), so the products were tested on or before then; recorded as 2025, the year of publication.
  30. LR-0030: One product (BM04) showed a different protein profile on SDS-PAGE (about 40.8 kDa), gave no signal on the anti-ACVR2B immunoblot and was identified as follistatin 344 by an anti-follistatin immunoblot. Identity only: an immunoblot shows which antibody-reactive protein is present; it does not measure purity or amount. PARTY-17 is this one anonymous supplier. Test date: The paper does not date the analyses. It was published in 2025 (Drug Testing and Analysis volume 17, issue 10; PubMed date May 1, 2025), so the products were tested on or before then; recorded as 2025, the year of publication.
  31. LR-0031: One product (BM14) showed no protein on SDS-PAGE with Coomassie stain (about 10 micrograms loaded per lane), so no ACE-031. The paper says mass spectrometry identified ipamorelin in it instead but does not show those data. Identity only: the gel shows that no protein was there; it does not identify or measure what the vial held. PARTY-18 is this one anonymous supplier. Test date: The paper does not date the analyses. It was published in 2025 (Drug Testing and Analysis volume 17, issue 10; PubMed date May 1, 2025), so the products were tested on or before then; recorded as 2025, the year of publication.
  32. LR-0032: An injection vial labelled hexarelin, obtained by the laboratory from the Anti-Doping Authority of the Netherlands together with vials labelled growth hormone and erythropoietin. Liquid chromatography with mass spectrometry found hexarelin, 'as indicated on their labels'. The paper reports identity only: no amount, purity or sterility for this vial. Where it was bought is unknown. The two vials labelled growth hormone in the same set held no active substance. Test date: The paper reports on 'different confiscated products and legally obtained nutritional supplements' that 'were analyzed in 2009'; recorded as 2009.
  33. LR-0033: Tablets sold over the counter as a nutritional supplement in Cyprus, labelled to contain GHRP-2 with three amino-acid ingredients. Liquid chromatography with tandem mass spectrometry, against a GHRP-2 reference standard, found GHRP-2 with the correct sequence at about 50 µg per tablet. The chemical structure printed on the label was wrong: it is not GHRP-2 or any other growth-hormone releasing peptide, and it was not in the tablets. The paper does not give the amount the label claimed, so the 50 µg cannot be judged against it, and it reports no purity. Test date: The paper reports on 'different confiscated products and legally obtained nutritional supplements' that 'were analyzed in 2009'; recorded as 2009.
  34. LR-0034: A drinking solution sold over the counter as a nutritional supplement in Cyprus, labelled to contain 'active GHRP-2 ester Z11'. Liquid chromatography with tandem mass spectrometry, against a GHRP-2 reference standard, found GHRP-2 itself, about 9 mg in one ampoule (one recommended daily serving). 'GHRP-2 ester Z11' is not a defined molecule, so whether the product held what its label named cannot be judged; the reading is inconclusive. No purity was reported. Test date: The paper reports on 'different confiscated products and legally obtained nutritional supplements' that 'were analyzed in 2009'; recorded as 2009.
  35. LR-0035: A glass vial of white powder for intranasal use, marked '50 mg Epitalon', seized by the Belgian health authorities. LC-MS/MS and high-resolution mass spectrometry found a peptide of the expected mass (391.1465 measured against 391.1466 for protonated AEDG); de novo sequencing and heat and acid treatment against synthetic reference peptides ruled out the isomer EADG and the isobaric TPCA, leaving AEDG. The paper reports identity only: it measured neither the amount of peptide nor its purity. Where the product was bought is not stated. Test date: The paper does not date the analyses. It says the laboratory received the two samples 'in the last year', and it was first published online on December 22, 2014; recorded as 2014.
  36. LR-0036: Sublingual tablets sold as a food supplement, marked 'EPITALON' and 3 mg of peptide per tablet, seized by the Belgian health authorities. LC-MS/MS and high-resolution mass spectrometry found the same peptide as in the vial (391.1467 measured against 391.1466 for protonated AEDG), confirmed as AEDG by the same reference-peptide tests. The 3 mg per tablet was not checked. The paper reports identity only: it measured neither the amount of peptide nor its purity. Where the product was bought is not stated. Test date: The paper does not date the analyses. It says the laboratory received the two samples 'in the last year', and it was first published online on December 22, 2014; recorded as 2014.
  37. LR-0037: The box and vials were intercepted by the Belgian medicines agency (FAMHP) at the end of 2014. An immunoblot against IGF-1, long R3 IGF-1 and R3 IGF-1 gave no signal for the sample; LC-MS/MS of the intact protein gave a mass of 6215.9 Da, matching none of them, and LC-MS/MS of its tryptic digest identified human epidermal growth factor, a cell-culture reagent. The paper gives no amount or purity. The label misspelled its own ingredient ('reconbinant'). Where the vial was bought is not stated. Test date: The paper says the preparation was submitted for analysis 'at the end of 2014'; recorded as 2014.
  38. LR-0038: A vial of powder labelled Melanotan II, seized by police with a vial of liquid meant to dissolve it. Retention time, UV spectrum and the doubly charged ion matched a purchased melanotan II reference standard on HPLC-DAD and UHPLC-TOF-MS. Where the vial was bought is not stated. Test date: The paper does not date the analysis. It was published in the November 2024 issue of the Journal of Forensic Sciences (online September 20, 2024), so the vial was analysed on or before then; recorded as 2024.
  39. LR-0039: The vial held 8.7 mg of powder, not the 10 mg on the label, and that powder was 30% melanotan II by HPLC-DAD against a reference-standard calibration curve: 2.61 mg of melanotan II, about a quarter of the labelled amount. No impurity peaks were seen; the authors suggest the rest was inorganic material such as buffers or salts. A dose measured by the label would hold about a quarter of what it claimed. Test date: The paper does not date the analysis. It was published in the November 2024 issue of the Journal of Forensic Sciences (online September 20, 2024), so the vial was analysed on or before then; recorded as 2024.
  40. LR-0040: Pens presented as Ozempic held insulin glulisine, a fast-acting insulin, with the excipients of an insulin pen and without those of genuine Ozempic; LC-HRMS, NMR and a polypeptide fingerprint agreed. People in Austria who injected these pens were admitted to hospital. Injecting insulin while expecting semaglutide can cause severe hypoglycaemia. The report counts no pens; its figures show two. Where the pens were bought is not stated. Test date: The report does not date the analyses. It was published by CVUA Karlsruhe on November 22, 2023, so the pens were analysed on or before then; recorded as November 2023.

What we searched for

For the compounds people buy most outside licensed pharmacies, we searched PubMed for published tests of products as sold. "None found" means none we could add under the rules above, on the date shown; the queries and the papers read and left out are listed so the search can be repeated.

Read and not added

Studies we screened that do not meet the rules yet. They are listed so that leaving them out is visible, not silent.

← Verify