There is a widely repeated notion that subcutaneous injection sites are interchangeable and that the standard set — abdomen, thigh, upper arm — applies to anything given under the skin. The approved labels do not support it.
Four peptide products with FDA labels are given subcutaneously. Each specifies a different set of permitted sites. This page reports what each states, and what each says about why. It is documentary: nothing here is an instruction, and administration decisions belong with a clinician.
The four lists
| Product | Molecule | Permitted sites per label | Frequency |
|---|---|---|---|
| WEGOVY | semaglutide | Abdomen, thigh or upper arm | Once weekly |
| ZEPBOUND | tirzepatide | Abdomen or thigh; back of the upper arm only if another person injects | Once weekly |
| EGRIFTA SV | tesamorelin | Abdomen only | Once daily |
| VYLEESI | bremelanotide | Abdomen or thigh | As needed, max 8 doses/month |
Four products. Same route. Four answers.
What the labels say about absorption — and it is the opposite of what the lists imply
The obvious hypothesis is that each list reflects where that molecule is absorbed adequately. Each label's own pharmacokinetics section rules that out.
The WEGOVY label reports that similar exposure was achieved with subcutaneous administration of semaglutide in the abdomen, thigh or upper arm. The ZEPBOUND label reports the same for tirzepatide across the same three regions. The VYLEESI label states that the site of subcutaneous administration — abdomen and thigh — had no significant effect on systemic exposure to bremelanotide.
So on three of the four labels, the manufacturer's own data says the region does not meaningfully change how much drug reaches the circulation. The site lists are narrower than the absorption data would require, and something other than absorption is setting them.
What is setting them, product by product
ZEPBOUND's upper-arm condition is about reach. Its label permits the back of the upper arm only when a second person performs the injection, while stating that upper-arm administration gives similar exposure. The back of one's own upper arm cannot be reached with the injecting hand. A physical constraint on self-administration has been written into the site list — which is why the same anatomical region is unconditional on the WEGOVY label, whose wording does not specify the back of the arm.
EGRIFTA SV's abdomen-only restriction is about local tissue, and the label quantifies it. Injection site reactions occurred in 25% of EGRIFTA-treated patients against 14% on placebo over the first 26 weeks of clinical trials — erythema, pruritus, pain, irritation and bruising. The label pairs that with rotation between different areas of the abdomen, and excludes scar tissue, bruises and the navel. It is also the only daily product in this set. A once-daily injection puts roughly seven times the number of punctures through the same tissue per week as a once-weekly one, and the label's structure follows that: one region, rotated within, with a stated reaction rate behind the instruction.
VYLEESI is delivered by autoinjector. The device fixes depth and delivery speed, which removes several of the variables a syringe leaves open, and its site list is correspondingly short.
Rotation is a tissue measure, not a dosing one
All four labels direct rotation. Only EGRIFTA SV states a reason, and that reason is reducing injection site reactions.
This is worth stating plainly because rotation is often described as though it affected how a dose behaves. On these labels it does not: the same documents report that the region made no meaningful difference to systemic exposure. What repeated injection into one spot affects is the tissue at that spot. That is the concern the labels are addressing.
What this means for compounds without labels
Nothing in the research-peptide market has an approved label, so nothing in it has a site list derived from its own clinical data. The conventions circulating on clinic and vendor pages for BPC-157, TB-500, CJC-1295 and the rest are conventions — the same class of claim as the dose figures catalogued on the dosage index, and gradeable the same way.
Three molecules blur the line and should not be allowed to. Bremelanotide is PT-141, elamipretide is SS-31, and tesamorelin appears in both markets. In each case the approved label describes a specific product — a specific formulation at a specific concentration in a specific device — and not a research vial of the same molecule. The site list on a label is part of that product's description, not a property of the molecule.
The four labels above make the point better than any argument could: four products, one route, four different lists, and the differences turn on the device, the frequency and the local tolerability of each particular product. That is not information a molecule carries.
What a label's site list is, precisely
It is a product-specific finding, and the four cases here show how product-specific. Absorption data said all three regions were equivalent for semaglutide and tirzepatide; the two labels still differ from each other on the upper arm, because one adds a condition about who is injecting. Tesamorelin is confined to one region against a measured 25% injection-site reaction rate. Bremelanotide's list is short partly because the product is an autoinjector.
Reading any one of those as "the sites for peptide injections" discards the reasoning that produced it.
Related on this site: how to reconstitute peptides covers the arithmetic between a vial and a syringe, syringes for peptides covers what a syringe unit is and is not, and bacteriostatic water covers the diluent those labels specify — with EGRIFTA SV a deliberate exception, since it is reconstituted with sterile water rather than bacteriostatic water, a distinction its label makes explicitly.