The short version

Plain English first. Everything below this section is the same material with the sources, the numbers and the caveats attached.

What AOD-9604 is

A sixteen amino acid piece of human growth hormone, closed by a disulfide bridge, with one deliberate change. It is the C-terminal tail of hGH — the region that was thought to carry the fat-mobilising activity — cut away from the part that makes you grow.

It differs from “HGH Fragment 176-191” by one oxygen atom

The two are sold side by side, often at similar prices, frequently with identical marketing copy. Native hGH(176–191) begins FLRIVQCRSVEGSCGF. AOD-9604 begins YLRIVQCRSVEGSCGF. Tyrosine is phenylalanine plus a hydroxyl group, so the formulae differ by exactly one oxygen — O₂₂ against O₂₃. That is the whole chemical difference.

Which produces an odd situation. The more-studied of the two failed its trials, and the less-studied one is trading on the reputation of those failed trials.

Why the tyrosine? The usual explanation is radioiodination convenience plus metabolic stability, and tyrosine is indeed the standard residue for iodine labelling. But we could not find a peer-reviewed paper stating that rationale, and neither could the one vendor honest enough to say so in print. The substitution is fact; the reason for it is convention.

Sourcing note on the chemistry block: PubChem and FDA's substance registry are JavaScript applications we could not render, and PubMed served a challenge page on every attempt. The formula, weight and CAS above rest on a commercial analytical-standards supplier rather than a primary registry. They are internally consistent with the sequence and with the one-oxygen relationship above, which is corroboration but not verification.

How it works

The design brief was to split growth hormone in two: keep the fat effect, drop the growth effect. Half of that is well established. It is not the half anyone buys it for.

So the molecule succeeded completely at not being growth hormone, and failed completely at burning fat. Stated that way it is not a selective drug. On the human evidence, it is an inactive one.

What the receptor experiment actually showed

The best-sourced mechanistic statement available says AOD-9604 “is not a high affinity antagonist of hGH binding to the GH receptor.” Read that carefully. It is a finding about competition — the peptide does not displace growth hormone from its receptor. It is not a demonstration that AOD-9604 acts through some other named receptor, and no such demonstration exists that we could find. No dissociation constant is reported.

The β3-adrenergic story runs backwards

Vendors say AOD-9604 activates β3-adrenergic receptors in fat tissue. The paper they are ultimately drawing on concludes the opposite. Heffernan and colleagues, 2001, in Endocrinology, verbatim:

The receptor is upregulated, and the chronic weight effect disappeared in knockout mice — so it is somehow involved. But the peptide is not acting as a β3 agonist, and the authors say so in their own conclusion. Note also that this is mouse β3-adrenergic pharmacology, which has a long and unhappy history of not transferring to humans; a generation of β3 agonists worked beautifully in rodents and failed as human weight-loss drugs. We found no human tissue or human receptor data at all.

A separate proposed mechanism — inhibition of acetyl-CoA carboxylase in liver and fat cells — is anti-lipogenic, meaning it blocks fat storage. Vendors routinely present it alongside lipolysis as though the two were the same claim. They are not.

Pharmacokinetics

Real numbers exist. They are unflattering, almost none of them are human, and the figure that circulates is off by an order of magnitude.

The route problem, which is the largest single fact on this page

The first two trials were intravenous. Every trial after that — including both phase 2b studies that decide this compound's fate — was oral tablets.

The market sells AOD-9604 as a subcutaneous injection at 300 to 500 mcg a day. That is a different route from the one that generated all the efficacy data, and as far as we can establish:

What the evidence shows

This is why the page exists. AOD-9604 is not under-studied — it is a compound whose question was asked properly and answered no.

All six randomised and placebo-controlled, all double-blind, run between 2001 and 2006 by Metabolic Pharmaceuticals of Melbourne. About 900 obese adults in total.

The number every vendor quotes

A 12-week trial in which subjects on 1 mg a day lost 2.6 kg against 0.8 kg on placebo — a placebo-adjusted 1.8 kg. That figure is real and it is where the entire marketing story comes from. Three things about it:

• The dose-response runs backwards. The arms were 1, 5, 10, 20 and 30 mg. The benefit was at 1 mg — the lowest dose tested, and the 10 mg group showed a smaller reduction. A thirty-fold dose range in which the smallest dose wins and a ten-times-larger dose does worse is not a pharmacological gradient.
• Six arms, about fifty per arm, no stated correction for multiple comparisons. With five active doses tested against placebo, something crossing p < 0.05 somewhere is close to expected before any drug effect at all.
• Even this figure is at two removes. It reaches us through a review article quoted in an FDA filing. We could not open the review, and the trial's efficacy analysis was never published.

The trial built to confirm it

METAOD006 was designed to chase exactly that signal. It dropped the entire upper dose range and tested 0.25, 0.5 and 1 mg — bracketing and centring on the winning dose — in 502 randomised subjects for twenty-four weeks, twice as long. That is a well-constructed confirmatory study.

The patent owner's own statement to the Australian stock exchange in 2013 says the same thing in softer words: the trials “proved that AOD9604 was profoundly safe” but showed no “clinically meaningful weight loss outcome.” The programme ended in February 2007.

Two things worth saying in fairness

The sponsor had an explanation, and it is not a silly one. Their account is that the final trial incorporated an intensive diet and exercise regime, and a strong lifestyle background compresses the difference between groups. That is a real methodological effect. It is also the argument every failed obesity drug makes, and it does not rescue a compound whose only positive signal sat at the bottom of an inverted dose curve.

The definitive efficacy data has never been peer-reviewed. The one published paper covering all six trials is explicitly scoped to safety and reports no weight-loss figures whatsoever. So the 536-subject, 24-week result that closes the question on this molecule exists only as a company statement and a line in a regulatory filing. Everything the market knows about whether AOD-9604 works — in either direction — is second-hand.

Where the fat-loss story actually comes from

Rodents. Two studies carry it:

• Obese Zucker rats, oral, 19 days. Weight gain reduced by more than half — 15.8 g gained against 35.6 g in controls. Note the endpoint: both groups got heavier. This is attenuated growth in a genetically obese rat strain, not fat loss, and vendors present it as the latter. We could not open the primary abstract; these figures come from a secondary record and should be treated accordingly.
• Obese mice, intraperitoneal, 14 days. Body weight and fat reduced; β3-adrenergic receptor expression restored toward lean levels. Discussed in the mechanism section, where the paper's conclusion turns out to undercut the marketing.

Neither rodent study used the subcutaneous route either. One was oral, one intraperitoneal. There is no subcutaneous efficacy study of AOD-9604 in any species that we could find.

The joint and cartilage claims

After obesity failed, the asset was repositioned. A 2015 study injected AOD-9604 into the knees of 32 rabbits with chemically induced osteoarthritis, alone or with hyaluronic acid, under ultrasound guidance. The combination gave lower cartilage degeneration scores and faster lameness recovery than saline. That is the entire basis for the joint-repair claims.

Rabbits, intra-articular, four groups. No human osteoarthritis trial of AOD-9604 could be confirmed — though we note the trial registries are JavaScript applications this environment could not query, so read that as “none found” rather than “none exists.” A professor involved in the Australian episode below put the repositioning bluntly: there was “no clinical evidence that it helped with tissue repair or had any other benefit in people.”

Reconstitution

Less sourced than we would like, and one structural feature that genuinely matters.

We could not find a manufacturer's quality-control insert for AOD-9604 comparable to the documentation that exists for some other compounds here. The one analytical supplier we could read gives “soluble in DMSO” and no aqueous figure at all — no mg/mL in water, in saline, or in anything else. There is no published stability study of AOD-9604 in bacteriostatic water. Everything else circulating is vendor procedure.

For the volumes people actually use:

Arithmetic only. The dose it serves is market convention, not a trial dose — see the protocols section, where the gap is unusually revealing.

We found no published survey of what is actually in internet-sold AOD-9604 vials that we were able to read. There is one relevant paper — a 2014 analysis of preparations seized by the Belgian authorities — and the publisher returned an error on every attempt, so we have its title and DOI and nothing else. We are not going to characterise findings we have not read. The honest position: unlike Melanotan II, where two regulators opened the vials and published the numbers, nobody has published an accessible purity survey of this compound. Buy against a lot-specific certificate; ours are on the certificate register.

Reported protocols

On most pages in this library, the protocols section says the circulating doses cannot be traced to a trial. Here the problem is the reverse, and it is worse: they can.

The most plausible reading is that the market dose was reverse-engineered from a 5 mg vial and a convenient injection volume, not from the literature. Twelve units on a U-100 syringe is a tidy number. It is not a finding.

Credit where it is due: at least one dosing site states plainly that its protocol “reflects the most commonly documented community approach” rather than evidence, and notes that the weight loss was “not replicated in later studies.” That is closer to honest than this category usually manages.

Storage

Lyophilised, sealed, cold, dark and dry. Our storage guide covers the general conventions.

The analytical-supplier guidance for the solid is 0 to 4 °C for days to weeks, −20 °C for months to years, dry and dark, with a stated shelf life over two years. For the reconstituted solution there is no sourced guidance at all — nothing measured, from any supplier, on stability in bacteriostatic water.

Reported effects

Unusually well documented, for the same reason everything else here is: it went through a real development programme. About 900 people took it under observation, and somebody wrote down what happened.

The summary from the published safety paper is that AOD-9604 was “indistinguishable from placebo.” That is a genuine finding, and it deserves context that cuts both ways: in the 12-week trial, 70% of subjects reported an adverse event during the placebo run-in, before anyone had taken anything. Most of the adverse-event burden in these trials is background noise, which is precisely why the drug could not be told apart from placebo on harms — or, as it turned out, on benefits.

The five cancers, handled carefully

The 12-week trial recorded five serious adverse events, all malignancies: three skin cancers in the 20 mg group, a breast cancer at 5 mg, a melanoma at 10 mg. The investigator judged none of them related to the drug.

FDA's own compounding entry notes that it has “identified serious adverse events that may be associated with AOD-9604, though causality is not clear.”

The asterisk on all of it

Chronic toxicology ran six months in rats and nine months in monkeys, with no adverse effect level found at the highest dose tested in either. No genotoxic activity across the standard battery. No published case reports of harm that we could find. On the human evidence this is one of the better-characterised compounds in the library.

Regulatory status

Approval: none, anywhere, for anything, ever. No FDA approval, no Orange Book listing, no pharmacopoeial monograph, no TGA registration, no Health Canada authorisation. Its developer terminated the programme in 2007 and the patent owner has said in writing that it has never been approved by any health authority in the world.

The FDA compounding position, which is widely misreported

AOD-9604 appears on FDA's list of bulk substances that may present significant safety risks — but not in Category 2. It sits under a separate heading: “nominated but withdrawn.” Somebody nominated it for compounding, then pulled the nomination.

FDA's entry, in full: compounded drugs containing AOD-9604 “may pose significant risk for immunogenicity for certain routes of administration”; the agency “has identified no, or only limited, safety-related information” and therefore “lacks sufficient information to know whether the drug would cause harm when administered to humans.”

Because its nomination was withdrawn rather than pending, AOD-9604 was not among the seven peptides heard at FDA's compounding advisory committee in July 2026, was not deferred, and is not queued. Several secondary sources list it as a Category 2 peptide awaiting review. It is not.

The GRAS claim

Vendors cite it constantly. Here is what it appears to be.

In June 2012 the patent owner announced that AOD-9604 had received GRAS status — generally recognised as safe — for use as a food ingredient at up to 1 mg a day. The one paper that mentions it describes the determination as “conditional on publication of pre-existing safety data” and names no organisation, no date and no regulatory body. We searched for an FDA GRAS notice and found none, and no FDA response of any kind. On the available evidence this is a self-affirmed determination: a company convenes an expert panel, the panel concludes GRAS, and notifying FDA is voluntary.

One footnote worth knowing. The GRAS determination was conditional on publishing the existing safety data — and the two open-access papers that constitute virtually the entire published AOD-9604 literature were written to satisfy that condition. Which is worth bearing in mind when reading them.

Canada

Get this one exactly right, because the obvious answer is wrong in both directions. Health Canada's 9 April 2026 advisory on injectable peptides names fourteen compounds. AOD-9604 is not one of them. But Health Canada did name AOD9604 — first in the list — in a seizure advisory dated 1 August 2025. It is an unauthorised drug in Canada either way. Anyone who says it is on the April 2026 list is wrong, and so is anyone who says Health Canada has never named it.

Sport

Prohibited at all times under S2, and named in the text of the list itself: “growth hormone fragments, e.g. AOD-9604 and hGH 176-191.” That wording has been in place since at least 2019. It would fail S0 as well — a discontinued development compound with no approval anywhere is the textbook case — but S0 only covers substances not addressed elsewhere, and this one is addressed by name.

The Australian football episode, since it is why most people have heard of it

During Essendon's 2012 season AOD-9604 was administered to players and appeared on their consent forms; the captain acknowledged taking it on television. At the time the compound was not named on the prohibited list, so the case for prohibition ran through S0, which is a materially weaker basis than a named listing. Adding to the confusion, it was simultaneously prescribable in Australia through a compounding pharmacy. WADA advised on 22 April 2013 that it fell under S0, and the anti-doping authority declined to pursue cases for use before that date.

Sentient Biogenix supplies AOD-9604 for laboratory research use only. It is not for human or veterinary use.

Questions we actually get

Does it work for fat loss?

Not in humans, on the best evidence available. A 12-week trial found 2.6 kg against 0.8 kg on placebo — at the lowest of five doses, with a worse result at ten times that dose. The confirmatory 24-week trial in 502 subjects, testing exactly that low-dose range, found no evidence of any difference from placebo at any dose. The developer shut the programme down.

Isn't it the same as HGH Fragment 176-191?

One oxygen atom different — a tyrosine where the native fragment has a phenylalanine. The difference matters for provenance rather than pharmacology: AOD-9604 had six human trials, and the native fragment has had none. The less-studied one is riding on the reputation of the more-studied one's failures.

Why is it sold by injection when the trials were oral?

No reason we could find. Every human efficacy trial after phase 1 used oral tablets. There has never been a human trial of subcutaneous AOD-9604 at any dose, for any duration, for any endpoint.

Is it safe?

On the human evidence, unusually so — about 900 subjects, up to 24 weeks, adverse events indistinguishable from placebo, no IGF-1 change, no glucose impairment, no antibodies, no published case reports of harm. With one large asterisk: all the chronic human data is oral, and FDA specifically flags immunogenicity risk for injection, which is how it is sold.

What's the half-life?

About three minutes after intravenous injection in pigs. Not the thirty minutes that circulates on dosing calculators, which cites nothing. No human pharmacokinetic study has ever been published.

Isn't it FDA GRAS?

It appears to hold a self-affirmed GRAS determination announced by its patent owner in June 2012, for food and supplement use at up to a milligram a day. We could locate no FDA GRAS notice and no FDA response. That is a food-safety conclusion about swallowing, reached by a panel the sponsor assembled — not a drug approval, not an efficacy finding, and not a sanction for injection.

Weren't the Essendon players banned for this?

No. They were suspended for thymosin beta-4. AOD-9604 cases for use before 22 April 2013 were never pursued.

Is it banned in sport?

Yes — S2, prohibited at all times, named explicitly in the list as a growth hormone fragment. The 2013 ambiguity that made it arguable is gone.

So why do you sell it?

Because it is a legitimate research compound with a real and unusually complete literature, and researchers have their own reasons for wanting it. What we will not do is tell you it burns fat. The trials were done properly and the answer was no, and you should know that before you buy anything from us.

References

1. Stier H, Vos E, Kenley D. Safety and tolerability of the hexadecapeptide AOD9604 in humans. Journal of Endocrinology and Metabolism 2013;3(1-2):7–15. DOI 10.4021/jem.v3i1-2.157. jofem.org — open access; the six-trial design table and every safety figure on this page. Read in full.
2. Moré MI, Kenley D. Safety and metabolism of AOD9604, a novel nutraceutical ingredient for improved metabolic health. Journal of Endocrinology and Metabolism 2014;4(3):64–77. DOI 10.14740/jem213w. jofem.org — open access; pharmacokinetics, metabolism, GH receptor binding, and the GRAS statement.
3. Heffernan M, Summers RJ, Thorburn A, et al. The effects of human GH and its lipolytic fragment (AOD9604) on lipid metabolism following chronic treatment in obese mice and β3-AR knock-out mice. Endocrinology 2001;142(12):5182–5189. DOI 10.1210/endo.142.12.8522. academic.oup.com — abstract read in full; the source of the β3 conclusion quoted above.
4. US Food and Drug Administration. Bulk drug substance nomination: AOD-9604 acetate. FDA-2024-N-4777-0002, attachment 4. downloads.regulations.gov — carries the phase 2b failure statement and the 2007 termination, quoting a March 2007 company filing we could not obtain directly.
5. US Food and Drug Administration. Certain bulk drug substances for use in compounding that may present significant safety risks. Updated 22 April 2026. fda.gov — AOD-9604 under “nominated but withdrawn”; the full agency entry quoted above.
6. US Food and Drug Administration. Pharmacy Compounding Advisory Committee, 23–24 July 2026. fda.gov — the agenda listing the seven peptides heard. AOD-9604 does not appear.
7. Calzada Limited. ASX announcement, 26 April 2013. announcements.asx.com.au — the patent owner's own account of the WADA advice of 22 April 2013, the June 2012 GRAS claim, the TGA non-approval statement and the February 2007 termination.
8. Health Canada. Unauthorized injectable peptide drugs seized and sold by Canada Peptide. Advisory, 1 August 2025. recalls-rappels.canada.ca — names AOD9604.
9. Health Canada. Think twice before injecting peptides bought online. Public advisory, 9 April 2026. recalls-rappels.canada.ca — the fourteen-peptide list. AOD-9604 is not on it.
10. Kwon DR, Park GY. Effect of intra-articular injection of AOD9604 with or without hyaluronic acid in rabbit osteoarthritis model. Annals of Clinical and Laboratory Science 2015;45(4):426–432. lateral-pharma.com — the 32-rabbit study behind the joint claims. Read in full.
11. Ng FM, Sun J, Sharma L, et al. Metabolic studies of a synthetic lipolytic domain (AOD9604) of human growth hormone. Hormone Research 2000;53(6):274–278. PMID 11146367. — the obese Zucker rat study. We could not open it; the figures quoted above come from a secondary record and are flagged as such in the text.
12. World Anti-Doping Agency. Prohibited List 2026, section S2.2.3. vada-testing.org — naming AOD-9604 explicitly. Wording confirmed identical in the 2019 edition.
13. Vanhee C, et al. Identification and characterization of peptide drugs in unknown pharmaceutical preparations seized by the Belgian authorities: case report on AOD9604. Drug Testing and Analysis 2014. DOI 10.1002/dta.1687. — we could not open it. Title, journal and DOI only. This is very likely the most valuable purity source that exists for this compound and we have not read a word of it.

Listed rather than hidden, what we could not check. The chemistry block rests on a single commercial supplier because PubChem and FDA's substance registry would not render and PubMed served a challenge page on every attempt. The 24-week failure statement reaches us through an FDA filing quoting a company filing we could not retrieve, and the 2.6 kg figure through a review article we could not open. We could not read the rat study, the Belgian purity analysis, or either of the two detection-method papers. No published human pharmacokinetics exists. No FDA GRAS notice was found, but we were unable to query the GRAS inventory directly, so read that as not found rather than not existing. Trial registries would not render, so “no human osteoarthritis trial found” is a search result, not a registry query. And the retrospective claim that WADA considered AOD-9604 prohibited from 2011 we have only at second hand.