AOD-9604: The GH Fragment With a Complicated Story

AOD-9604 is a peptide with an identity crisis. It was designed to be "growth hormone's fat-loss half" — a short C-terminal fragment that would burn fat without the growth, without the IGF-1 bump, without the diabetogenic side. Pharmacologically, that's a beautiful idea. Clinically, the Phase IIb trial in 2007 failed the primary endpoint. The drug-development track basically ended. And then it quietly got FDA GRAS status as a food ingredient in 2020, the cartilage-repair research picked up, and it never really left the peptide forums.

Most people don't run AOD-9604 expecting miracles. They run it because it's one of the cleaner-tolerated compounds in the fat-metabolism space, and because it stacks with things.

What it is, in one paragraph

AOD-9604 is a 16-amino-acid peptide corresponding to the C-terminal region of human growth hormone (residues 177–191) with an added N-terminal tyrosine. It keeps the part of GH that activates lipolysis — adipose fat breakdown — while missing the parts that cause GH's other effects (growth, IGF-1 elevation, insulin resistance). No GH receptor dimerization, no IGF-1 bump, no glucose trouble. The disulfide loop between its two cysteines is what makes it biologically active.

Dosing: what people actually do

Sub-Q, morning, empty stomach.

• Starting: 300 mcg once daily
• Standard: 500 mcg once daily
• Some users: 300 mcg AM + 300 mcg pre-workout (split dose)

Cycles run 8–12 weeks, sometimes up to 16. AOD is one of the compounds people run for months because the side-effect bill is so low, and because the effect is subtle enough that a short trial doesn't tell you much.

"I ran AOD for ten weeks alongside a cut. Did it do something? Probably. Did it do something I couldn't have gotten from being more patient with the diet? Hard to say. But it was the only peptide in the stack that never gave me any issue at all." — forum user

What it pairs with

This is where AOD actually shines — it's a supporting compound.

• CJC-1295 / Ipamorelin — the classic GH-secretagogue stack. AOD adds the C-terminal fragment's lipolytic action on top of pulsatile endogenous GH
• Semaglutide / Tirzepatide / Retatrutide — GLP-1 handles appetite, AOD works on adipose fat mobilization. Overlap is minimal, both act differently
• 5-Amino-1MQ — metabolic stack, adipocyte-level
• Tesamorelin — GHRH analog, FDA-approved for visceral fat; some users run them together for body recomp
• Cardio and caloric deficit — AOD without those isn't going to do much

Red flags and side effects

AOD's safety profile is one of the cleaner ones in the peptide space:

• Injection-site reactions — rare, mild
• Mild fatigue or headache — uncommon
• No IGF-1 elevation — genuinely clean on the GH-axis labs
• No insulin sensitivity issues
• GRAS status — unusual, and genuinely suggestive that the compound is very well-tolerated

The 2012–2013 AFL doping scandal in Australia put AOD on the WADA prohibited list. Competitive athletes need to know that.

The honest limits

• The big Phase IIb trial failed the primary endpoint. That's the central caveat. The trial didn't say AOD does nothing — it said it didn't beat placebo on weight loss over 24 weeks in the population tested
• Forum opinion is divided on whether the trial was underpowered/mistargeted or whether the compound simply isn't strong enough to drive weight loss on its own. The body of evidence leans toward: useful in a stack, not standalone
• Human PK data is thin. Rodent and in vitro data is more robust
• The cartilage-repair story is interesting but still early. If you're running AOD for joints, you're in experimental territory
• Product quality ranges wildly. The compound is popular enough that low-grade supply is everywhere. COA matters

Where to go next

• For the full mechanism, the Phase IIb context, and the GRAS/cartilage pivot, see the Pepperpedia AOD-9604 entry.
• For real stack logs — especially the CJC/Ipa + AOD combo — the Optimization forum is where people track what's working.
• For reconstitution math on a 5 mg vial, PepperCalc handles it.
• Comparing metabolic peptides? See 5-Amino-1MQ and IGF-1 LR3.