"Are peptides safe" is one of the most-searched questions in this field, and it cannot be answered as asked. The word covers three kinds of product that share chemistry and almost nothing else: medicines that went through FDA approval, pharmacy-compounded preparations, and powders sold online for "research use only". The checking behind each is different, so the safety record behind each is different.
This page sets out what each record holds and links to the per-compound pages rather than repeating them. Nothing here is a recommendation. Anyone deciding about any peptide should make that decision with a clinician who knows their history.
GLP-1 drugs are peptides, but their safety record — the largest of any class here — is covered by GLP1 Ledger's safety page and is not repeated on this one.
The short answer, by product type
| FDA-approved peptide | Compounded peptide | Research-use peptide | |
|---|---|---|---|
| Examples | Tesamorelin (Egrifta), bremelanotide (Vyleesi), teriparatide (Forteo), octreotide | Sermorelin, gonadorelin, glutathione drips | BPC-157, TB-500, CJC-1295, MOTS-c |
| Tested in controlled human trials | Yes, for the labelled use | Only if the molecule once was, and not the pharmacy's product | Rarely or never |
| Known adverse-reaction rates | Printed on the label | Borrowed from old trials, if any | None |
| Who checks identity and purity | FDA, under the approved application | State board; FDA for outsourcing facilities | No regulator |
| FDA's stated concerns | Label warnings | 20 peptide ingredients named on its compounding-risk page | Warning letters to sellers |
How many peptides are approved, and which, is on our FDA-approved peptides census.
Approved peptides: measured risks, on paper
An approval does not mean "safe". It means FDA judged that, for one labelled use, the measured benefits outweighed the measured risks — and the label states those risks. Three examples from labels this site has read:
- Tesamorelin (Egrifta WR, label revised 3/2025). Among patients treated for 26 weeks, 47% had IGF-1 more than 2 standard deviations above normal; anti-tesamorelin antibodies were found in 50%. The label states that "long-term cardiovascular safety … has not been established" and contraindicates it in pregnancy. The full record is on our tesamorelin side effects page.
- Injection-site reactions. Across ten approved peptide labels, rates run from under 1% to 58% (Tymlos, with daily checks), and placebo injections caused reactions in up to 28% — so part of the local reaction is the injection, not the peptide. Detail on our injection site reactions page.
- Allergy. The Increlex (mecasermin) label reports post-marketing allergic reactions from local injection-site reactions to anaphylaxis requiring hospitalisation.
The advantage of the approved class is not that its risks are small. It is that they are known, counted and printed, and that the product in the box is the product that was tested.
Compounded peptides: what FDA has said
A compounding pharmacy makes a preparation for a prescription. When the molecule was once approved — sermorelin and gonadorelin are both discontinued, not withdrawn for safety — the old trials are the only human data, and they did not test the pharmacy's product. When the molecule was never approved, there is no trial record at all.
FDA maintains a page of bulk drug substances that may present significant safety risks in compounding (content current as of 22 April 2026). Counted on 6 October 2026, it names 20 peptides or peptide fragments: 17 in its safety-risk table — AOD-9604, BPC-157, LL-37, CJC-1295, dihexa, DSIP (emideltide), epitalon, injectable GHK-Cu, ipamorelin, KPV, PEG-MGF, melanotan II, MOTS-c, selank, semax, thymosin alpha-1 and the TB-500 fragment — plus GHRP-2, GHRP-6 and kisspeptin-10 in its category-2 table. The reasons repeat across the entries:
- Immunogenicity — the body may make antibodies against the peptide, especially by injection.
- Peptide-related impurities — synthesis by-products, truncated or altered chains, and aggregation.
- Characterisation — some contain unnatural amino acids, which FDA says "adds to the complexity of peptide characterization."
- No or limited safety information for the routes proposed.
For a few entries FDA goes further. Of GHRP-2 it states it is aware of reports of serious adverse events including "increased insulin requirement to maintain the blood glucose level, death of critically ill study subjects, infection and pancreatitis, though causality has not been established." Its reasoning compound by compound is gathered on our peptide side effects hub.
In July 2026 FDA's staff advised against adding seven of these peptides to the list compounders may use, while its advisory committee voted for six of them. That disagreement, with sources, is on our peptide therapy page.
The ingredient can be the hazard
The clearest documented compounding harm involving a peptide in 2026 had nothing to do with the peptide's pharmacology. On 27 August 2026 FDA reported at least 30 patients with adverse events after intravenous glutathione — a tripeptide — compounded by different pharmacies from one lot of powder "labeled as dietary supplement grade, which is not an appropriate ingredient to make injectable drugs." The likely culprit was endotoxin, a bacterial residue that survives sterilisation. FDA had warned about the same failure in 2019. The account is on our glutathione injection side effects page.
That case is the general point in one example: for an injected product, the grade, sterility and endotoxin level of what is in the vial can matter as much as the molecule.
Research-use peptides: what nobody has measured
A "research use only" label means the product is not sold for people, so nobody is required to show it is safe for people. What that leaves:
- No safety trial of the product as sold. For most research-market names, the human record is a handful of small studies or none; per-compound counts are on our side-effect pages, for example BPC-157, TB-500 and MOTS-c.
- No regulator checking the vial. Identity, purity, sterility and endotoxin are whatever the seller and its certificate say. How to read a certificate, and what it cannot prove, is on our COA guide; why a colour-reaction kit cannot tell BPC-157, TB-500 and GHK-Cu apart is on our peptide test kits page.
- Enforcement rather than oversight. FDA warning letters dated 10 December 2024 to 24 August 2026 went to at least 15 US-facing research-peptide sellers for marketing unapproved new drugs. Our warning-letter tracker lists each one.
- Adverse-event reports that mostly point elsewhere. In FDA's adverse-event database, many reports naming a research peptide list it as one of several products taken, which makes the count hard to read in either direction — the problem is explained on our side effects hub.
What "research use only" does and does not permit is on our research-use-only page.
Four questions that sort a product
These are the questions the record can answer for any peptide product. They are not a substitute for a clinician.
- Is the product itself FDA-approved, or only the molecule — or neither? The census answers this.
- Who made it, and under whose inspection? An approved manufacturer, a licensed pharmacy or outsourcing facility, or an unregulated seller.
- What human data exist for this molecule, by this route, at this amount? Our per-compound dosage and side-effect pages count them.
- Is there an independent test of this lot for identity, purity and, for injectables, endotoxin? Our peptide testing page explains what each test shows.
What is not established
- Long-term safety in people for almost every research-market peptide.
- Whether compounded versions of discontinued peptides match the approved product's identity and potency.
- What sustained IGF-1 elevation from GH-axis peptides means for cancer risk over years; the tesamorelin label states that "the effects of prolonged elevations in IGF-1 levels are unknown" and asks for IGF-1 to be monitored.
- The real rate of harm from research peptides: adverse events go unreported when the product was not sold for human use.
Peptide Lexicon's are peptides good for you page counts the same field from the other side — where benefits have been measured.
Sources and dates
- Drugs@FDA via openFDA (dataset updated 2026-10-02), read 2026-10-06.
- FDA, Certain bulk drug substances for use in compounding that may present significant safety risks, content current as of 2026-04-22, read and counted 2026-10-06.
- EGRIFTA WR (tesamorelin) prescribing information, revised 3/2025, accessdata.fda.gov, read 2026-10-06.
- FDA glutathione alert of 2026-08-27, as read and cited on our glutathione injection side effects page.
- FDA warning letters, as listed with links on our warning-letter tracker.