What “compounding” actually means — and why US compounded peptides are a category of their own

Most medicines come off a factory line, sealed in identical packaging, in batches of millions. A compounded medicine is different. It is mixed, in small quantities, by a licensed pharmacist — usually because the off-the-shelf version doesn’t exist, doesn’t come in the right strength, or contains an ingredient a specific patient can’t tolerate. Compounding is the part of pharmacy that still resembles cooking from scratch, while ordinary pharmaceutical manufacturing resembles a packaged-food plant.

US compounded peptides are short chains of amino acids — the small building blocks that cells use to make proteins — prepared inside this compounding framework. They are not supplements, and they are not FDA-approved drugs. They occupy a third category defined by federal law, written into the books in 2013 after a deadly outbreak of fungal meningitis traced to a Massachusetts compounding pharmacy. Congress responded by passing the Drug Quality and Security Act, which split compounding into two legal flavors — 503A and 503B — each with its own rules.

The third regulatory category — neither supplement nor approved drug

That regulatory in-between is the thing to understand. These compounds aren’t bootleg, but they aren’t approved either. They sit in a corridor that exists because Congress wanted pharmacies to be able to fill specialized needs without forcing every preparation through the years-long process of new-drug approval — and the rules of that corridor are what this article is about.

503A and 503B: the two kinds of US compounding pharmacy, explained without legalese

The 2013 law split compounding pharmacies into two groups. The split is about volume, scrutiny, and whether a specific patient has to be involved.

503A — the traditional compounding pharmacy

A 503A pharmacy is what most people picture when they hear “compounding.” A doctor writes a prescription for a specific patient who needs a custom preparation; the pharmacy makes it. These pharmacies are licensed and inspected by their state’s board of pharmacy, not by the FDA directly. They must follow the United States Pharmacopeia compounding chapters — USP 795 for non-sterile preparations, USP 797 for sterile ones — which lay out how the work must be done and tested.

The 503A rules add up to a single sentence: you can compound, but only for a specific patient who has a prescription, and you have to follow the recipe and quality rules.

503B — the outsourcing facility

A 503B is different. It can compound larger batches without a patient-specific prescription, which means hospitals and clinics can order pre-made batches the way they would order any other medicine. The trade-off is that 503B facilities register with the FDA and are inspected the same way conventional drug factories are, under the FDA’s Current Good Manufacturing Practice (cGMP) regulations. They report adverse events to the FDA, and the agency can shut them down.

Here is the everyday version. A 503A is your independent corner bakery making one wedding cake to order. A 503B is a regional commissary baking five thousand identical cakes a week for restaurants — same trade, much heavier oversight. Both produce real cakes; the inspections are not the same.

What both have in common

Whatever the category, US compounding pharmacies are required to start from raw materials that meet pharmaceutical-grade standards, to test the finished preparation, to document every batch with a Certificate of Analysis (a lot-specific lab report showing what is actually in the vial), and to follow sterility rules for anything injectable. None of that is true of substances entering the country as “research chemicals” outside the pharmacy framework — and that distinction is what the rest of this article is about.

Why peptides are unusually hard to make well

A peptide is a chain of amino acids. If amino acids are letters, then peptides are short words and proteins are long sentences. A peptide drug works only if the letters are in the right order, the chain is the right length, the molecule is folded the right way, and nothing else is hitching a ride in the vial.

That last part — “nothing else is hitching a ride” — is where most quality failures happen. Synthesizing a peptide in a lab almost always produces a small percentage of misspellings: chains that are one letter short (called truncated peptides), chains where one amino acid got swapped, residual chemical reagents from the manufacturing process, and leftover bits of solvent. A pharmaceutical-grade peptide isn’t just “the right molecule” — it’s a vial where 98% or more of what’s in there is the right molecule, and the remaining 2% has been measured and accounted for.

How researchers actually check — the role of HPLC

Researchers use a technique called high-performance liquid chromatography (HPLC) to check. The basic idea is straightforward: pump the sample through a long, packed tube, and different molecules come out the other end at different times based on how they interact with the packing material. Plot what comes out and you get a chromatogram — a graph with a tall peak for the peptide you want and smaller peaks for everything else. The areas under those peaks tell you the percentages. A 2023 paper in Pharmaceutical Research on reference standards for synthetic peptide therapeutics describes the full analytical battery — HPLC, mass spectrometry, NMR — used by regulators and manufacturers to confirm that a peptide is what its label says it is.

Why peptides break easily — and why that matters

That sounds like overkill until you remember why it exists. Peptides are fragile in ways small-molecule drugs aren’t. They degrade when exposed to heat, light, and moisture. Their three-dimensional shape — the folding that lets them fit into a cell’s receptor like a key fits a lock — can shift under bad storage conditions, and a misshapen peptide may not work at all. Some impurities, particularly truncated chains, can also trigger immune reactions, because the body’s immune system is exquisitely sensitive to slight differences in protein structure. This is why injectable peptide preparations get tested for sterility and for endotoxins (bacterial cell-wall fragments that can cause fever) on top of identity and purity.

Put it together and the picture is this: a peptide that looks identical in two vials may, under HPLC, turn out to be 99% pure in one and 84% pure in the other, with a meaningfully different mix of impurities — and a non-scientist holding either vial would have no way to tell. The testing infrastructure compounded peptides operate under exists precisely because the molecule itself doesn’t volunteer its quality.

The FDA bulk substances list and how it’s changing in 2026

Even within the 503A framework, a pharmacy can’t compound from just any ingredient. The FDA maintains a working list — the 503A Bulk Drug Substances list — that sorts active ingredients into three categories.

• Category 1 — substances the FDA has evaluated and considers eligible for compounding under 503A.
• Category 2 — substances the FDA has flagged with safety concerns; not eligible for routine 503A compounding.
• Category 3 — substances under review, with not enough information yet for a determination.

For years the most-discussed compounded peptides — including BPC-157 and TB-500 — sat in Category 2, which effectively shut the door on 503A compounding for human use. The FDA cited concerns about immunogenicity (the immune-reaction risk mentioned earlier) and impurity profiles. That was the regulatory baseline most peptide-compounding coverage was written against.

The April 2026 reshuffle

In April 2026 the picture changed. Following the withdrawal of certain nominations, the FDA removed twelve peptides from Category 2: BPC-157, TB-500, LL-37, DiHexa, DSIP (Delta Sleep-Inducing Peptide), Epitalon, injectable GHK-Cu, KPV, PEG-MGF, Melanotan II, MOTs-C, and Semax. Industry observers described it as the most significant shift in years.

Here’s the catch — and it is a real one. Removal from Category 2 didn’t mean those peptides moved to Category 1. They simply left the “do not compound” list without joining the “approved to compound” list. The result, as one regulatory analysis from a US compounding pharmacy network put it, is a gray zone: neither explicitly prohibited nor authorized. Pharmacies have to make judgment calls; the FDA has signaled it will reconsider these substances individually.

What the Pharmacy Compounding Advisory Committee is doing in July 2026

The FDA’s Pharmacy Compounding Advisory Committee is meeting on July 23 and 24, 2026, specifically to review whether several of these peptides should be added to the 503A Bulks List. The first day’s agenda covers BPC-157, KPV, TB-500, and MOTs-C; the second day covers DSIP (Emideltide) and Semax. Other peptides, including LL-37 (cathelicidin), DiHexa, PEG-MGF, GHK-Cu, and Melanotan II, are expected to be reviewed by early 2027.

That is the most important thing to know about this category as of mid-2026: the regulatory ground is moving. Anyone writing about, researching, or watching this space should expect the list of legally compoundable peptides to look different a year from now than it does today.

The import landscape: what is documented and what is just rumored

One claim travels widely in the peptide-sourcing conversation: that hundreds of millions of dollars of peptide material enter the US each year, mostly from China, classified as “research chemicals” to dodge the FDA’s drug-import requirements. The general phenomenon is real and documented by the FDA itself — the agency has issued repeated warnings and import alerts on bulk peptide shipments. Specific dollar figures floating in industry trade press, however, are usually un-sourced and should be treated cautiously. The honest summary is that a meaningful volume moves through that channel; the precise total is not well established.

What’s documented vs. what’s headline material

What is established is what falls away when a peptide is imported and resold outside the pharmacy framework. There is no lot-specific Certificate of Analysis traceable to a US pharmacy. There is no required HPLC test on the finished, packaged vial. There is no sterility testing requirement for injectables. There is no adverse event reporting infrastructure, no formal recall mechanism, and no FDA inspection of the synthesis facility. Whether any of those things actually happened is then a matter of trust in the seller’s voluntary disclosures.

Independent testing programs that have purchased samples from gray-market vendors and run them through accredited labs have repeatedly found discrepancies — lower purity than the labeled claim, undeclared impurities, and in a small number of cases the wrong peptide sequence in the vial. Those findings are not new; reporting in venues including STAT and The New York Times has documented them for years. The findings are also why the FDA’s compounding framework exists in the first place: not because regulators believe every overseas vial is bad, but because there is no way for a downstream user — researcher or otherwise — to verify quality without lab equipment most people don’t have.

What the research can’t tell us yet

What studies and regulators have actually documented

Six findings that come from peer-reviewed literature and FDA documents — not from marketing copy — and that together define the current state of evidence on US compounded peptide quality.

None of these findings, individually, settles the question of whether a particular peptide does what proponents claim it does in humans. Together, they describe the analytical and regulatory floor that defines what “pharmaceutical-grade” actually means in this space.

How to read the evidence on US compounded peptides

Reading research well isn’t the same as making purchasing decisions. It’s the skill of holding a finding at the right confidence level — taking it seriously without overclaiming. Five factors that help.

1. Lab dish vs. animal vs. human

A finding in a petri dish (in vitro) tells you what a molecule can do under controlled chemistry. An animal study tells you whether the molecule does something in a living mammal. A human trial tells you whether it does that thing in people. Each step adds enormous complexity. Many peptide findings in the popular discussion are at the cell or rodent level; treating them as if they were human trial results overstates what is known.

2. Sample size and replication

A single small study, however well-designed, is a starting point — not a settled question. Findings get more reliable when independent groups in different labs find the same thing. For these peptides specifically, several headline claims rest on a small number of studies that have not been independently replicated.

3. Who funded it

Funding doesn’t automatically corrupt research, but it is a useful filter. Studies funded by a substance’s manufacturer or by clinics that sell it are not invalid, but they are worth weighing differently than independent academic work. Reputable studies disclose funding sources in the paper.

4. The regulatory snapshot at the time of writing

An article written in 2023 about BPC-157 was written when that peptide was Category 2 — effectively un-compoundable. An article written in May 2026 was written after Category 2 removal but before the PCAC review. Both can be technically accurate at the time of writing and outdated within months. Date-stamp anything you read in this space.

5. Distinguishing “what the molecule does” from “what people sell it for”

This is the single most useful filter when reading about compounded peptides. Researchers study what a peptide does in their model system. Clinics describe what they hope it will do for their patients. Marketing describes what people would pay for it to do. These are three different conversations, and they often use the same words.

Side-by-side: 503A, 503B, and unregulated supply

This table compares the regulatory frameworks themselves, in plain language. It is not a comparison of which is more effective for any health purpose — it describes what each framework requires.

The point of the table isn’t that 503A and 503B are interchangeable, or that they make a peptide effective; the point is that each row describes a real, enforceable requirement that has no counterpart in the unregulated supply chain.

Open research questions on US compounded peptides

Some of the most interesting work in this field is happening on questions that are still wide open. Four research directions worth watching.