FDA maintains one table of bulk drug substances it has identified as presenting significant safety risks in compounding. It is cited constantly in writing about peptides, usually second-hand and almost never in full. This page indexes it: every substance, which statutory section it falls under, and the date it was added, read directly from FDA's page on 13 September 2026.
What the two sections mean, and why the same substance can carry different answers under each, is set out in our explainer on 503A and 503B compounding. This page does not repeat that argument. It publishes the record.
The whole category 2 table
Fourteen substances, as the page stood when read. FDA's page content is marked current as of 22 April 2026.
| # | Substance | Section | Date added |
|---|---|---|---|
| 1 | Domperidone | 503A | 27 October 2015 |
| 2 | Germanium sesquioxide | 503A | 9 June 2016 |
| 3 | Quinacrine hydrochloride (intrauterine) | 503A | 9 June 2016 |
| 4 | Cesium chloride | 503A | 30 July 2018 |
| 5 | Quinacrine hydrochloride (intrauterine) | 503B | 7 June 2021 |
| 6 | Chloral hydrate | 503B | 5 October 2022 |
| 7 | Edetate disodium (except ophthalmic) | 503B | 12 October 2022 |
| 8 | Neomycin sulfate (parenteral, non-ophthalmic/otic) | 503B | 12 October 2022 |
| 9 | Ibutamoren mesylate | 503B | 29 December 2022 |
| 10 | Diethylstilbestrol | 503B | 21 August 2023 |
| 11 | GHRP-2 (injectable/nasal) | 503B | 29 September 2023 |
| 12 | GHRP-6 | 503B | 29 September 2023 |
| 13 | Ipamorelin acetate | 503B | 29 September 2023 |
| 14 | Kisspeptin-10 | 503A | 29 September 2023 |
| 15 | Ibutamoren mesylate | 503A | 29 September 2023 |
| 16 | Tranilast | 503B | 29 September 2023 |
Sixteen rows, fourteen substances: quinacrine and ibutamoren each appear twice, once under each section. Peptide-class entries are in bold.
FDA's stated grounds vary by substance and are summarised on the page itself. For the peptide entries they cluster on immunogenicity — the risk that an injected peptide provokes an immune response — together with insufficient safety data, and for the secretagogues, endocrine effects. Both grounds are prospective rather than retrospective, which they had to be: when we counted the reports FDA actually holds for these compounds, ipamorelin returned eleven and kisspeptin none, which is far too little to reason from. The empty adverse-event column is not for want of clinical study — the published human record for kisspeptin runs to dozens of supervised infusion protocols with doses stated to two decimal places. For the small molecules they are conventional and severe: cardiac arrhythmia and sudden death for domperidone, nephrotoxicity for germanium sesquioxide, carcinogenicity for diethylstilbestrol, confusion between two similarly named chelating agents for edetate disodium.
Three things the dates say
Nearly half the list arrived on one day, and that day was about peptides. Six of the fourteen substances carry the date 29 September 2023. Five of those six are the peptide-class entries; the sixth is tranilast. Before that day the table held eight substances and one peptide. After it, it held fourteen and five.
The peptide entries are recent, and the list is not. Domperidone has been on it since October 2015 — almost eight years before any peptide joined it. A reader arriving at this table from a peptide article will naturally read it as a peptide list. It is not one, and never was: nine of the fourteen entries are conventional small molecules, several with decades of clinical history behind the concern.
Nothing gets added to both sections at once. The two substances that carry entries under both — ibutamoren mesylate and quinacrine hydrochloride — acquired them nine months and just under five years apart respectively. That is the clearest available evidence that these are two processes rather than one: a single decision applied to both sections would produce a single date, and it never has.
Four of the five peptides are the same kind of thing
Read as a class rather than a list, the September 2023 action narrows further. GHRP-2 and GHRP-6 are growth hormone releasing peptides. Ipamorelin is a growth hormone secretagogue. Ibutamoren — sold in the research market as MK-677, and not a peptide at all but a spiroindane small molecule — is an orally active agonist at the same receptor. Four of the five peptide-class entries act on one axis.
Kisspeptin-10 is the exception, and it is also the only peptide entry placed under 503A rather than 503B.
That matters for anyone reading the table to work out what FDA thinks about peptides generally. On this record the agency has not acted against peptides as a category. It has acted against growth hormone secretagogues, plus one reproductive-axis peptide, in a single 2023 action, under whichever section each was nominated.
The withdrawn list is the one the research market should read
Lower on the same FDA page sits a separate section for substances nominated for use in compounding whose nominations were later withdrawn. Sixteen substances are listed, and this is the finding worth the page:
| Substance | Also known as |
|---|---|
| AOD-9604 | hGH fragment 176-191 |
| BPC-157 | body protection compound 157 |
| Cathelicidin LL-37 | LL-37 |
| CJC-1295 | — |
| Dihexa acetate | — |
| Emideltide | DSIP, delta sleep-inducing peptide |
| Epitalon | epithalon, AEDG |
| GHK-Cu (injectable routes) | copper tripeptide-1 |
| KPV | α-MSH 11-13 |
| Mechano growth factor pegylated | PEG-MGF |
| Melanotan II | — |
| MOTS-c | — |
| Selank acetate | TP-7 |
| Semax | heptapeptide |
| Thymosin alpha 1 | Tα1, thymalfasin |
| Thymosin beta-4 fragment | TB-500 |
Every one of the sixteen is a peptide. Where category 2 is five peptides out of fourteen entries, the withdrawn section is sixteen out of sixteen — and the substances in it are, close to item for item, the catalogue a research supplier sells. Thirteen of the sixteen carry a dosage reference on this site, which is a reasonable proxy for how much search demand each one draws; the three without one are dihexa acetate, PEG-MGF and cathelicidin LL-37.
Seven of the sixteen are under active reconsideration. On 23-24 July 2026 FDA's Pharmacy Compounding Advisory Committee considered BPC-157, KPV, TB-500, MOTS-c, emideltide (DSIP), semax and epitalon for the 503A bulks list — the list that would permit compounding from them. It recommended six and voted against emideltide, with semax passing 8-5, epitalon 7-5 with one abstention, TB-500 8-6 against FDA staff's own position, and emideltide failing 6-7 with one abstention. Those votes are advisory, FDA has not acted on them, and none of them changes the table above. What they do establish is that the withdrawn section is not a closed record. The vote and its limits are covered on our page on cognitive peptides.
A withdrawn nomination is a procedural status and nothing more. It means the process that would have produced a listing did not run to completion. It is not a safety finding, it is not a clearance, and it is misreported in both directions — sellers cite the absence of a restriction as though FDA had looked and approved, and critics cite the withdrawal as though FDA had looked and refused. Neither happened. What we can say about the regulatory position of these substances is set out in our page on what the rules actually say about peptides.
What this table cannot tell you
It says nothing about a vial bought from a research supplier. Both sections describe compounding — a pharmacist, physician or registered facility making a drug for human use under defined conditions. Material shipped under the caution labelling of 21 CFR 312.160 sits elsewhere in the framework entirely, which is the distinction FDA's own enforcement correspondence turns on in the letters we track on the FDA warning letter tracker.
It says nothing about what is in any particular vial. A substance's listing status is a fact about the substance; what a seller shipped is a question answered by testing it, which is why a certificate of analysis is read the way our guide to reading a COA describes and a listing status is not a substitute for one.
And it says nothing about legality for any specific person, product or transaction. That is a legal question, and this is a publication rather than a law firm.
How to check this yourself
The table lives on one FDA page and takes about two minutes to verify. Open Certain Bulk Drug Substances for Use in Compounding That May Present Significant Safety Risks on fda.gov, count the rows, and check the "Content current as of" date at the foot of the page against the date on this one. If they differ, FDA's page is right and this one is stale — we re-check it on the same monthly cycle as the warning-letter tracker.
The one thing worth doing while you are there is reading which section each entry sits under, because that column is the one most often dropped when the table is quoted second-hand, and dropping it is what turns "restricted under 503B" into "restricted".
Related on this site
- 503A vs 503B — what each section requires, and why they produce different answers about the same substance
- What the rules actually say about peptides — 21 CFR 312.160 and the research-use-only position
- FDA warning letters to peptide sellers — the enforcement record, one row per letter
- How to read a certificate of analysis — what a COA commits a seller to, and what it does not
Sources and dates
- FDA, Certain Bulk Drug Substances for Use in Compounding That May Present Significant Safety Risks. Page content current as of 22 April 2026; read 13 September 2026. Both the category 2 table and the nominated-but-withdrawn section were read from this page.
- 21 U.S.C. § 353a (FD&C Act section 503A) and 21 U.S.C. § 353b (section 503B), for the statutory framework the categories operate under. Read 13 September 2026.
- Substance identities and synonyms cross-checked against PubChem records, read 13 September 2026.
This page is journalism about a public regulatory record. It is not legal advice, and it cannot tell any reader where a particular product or transaction sits.