"Pharmaceutical grade" isn't a legal term. Here's what is.

The short answer: neither "research grade" nor "pharmaceutical grade" is a category in US drug law. They're marketing words.

What the law does define is how a drug got made and who checked it. An FDA-approved drug, a compounded drug and a "research use only" vial sit in three very different rule books. The difference that matters most isn't purity. It's whether anyone was required to test the vial for sterility and endotoxin before it shipped.

Status as of October 21, 2026: no research peptide sold as "research use only" is approved for human use. This post explains labels and rules. It is not medical or legal advice and contains no dosing.

Where I get mine: the vendors I use, and how I vetted them.

Not in drug law. I couldn't find "pharmaceutical grade" defined anywhere in the FD&C Act or FDA's drug regulations, and the vendors using it never cite one.

The closest formal definition comes from animal research, of all places. Federal lab-animal oversight asks researchers to use pharmaceutical-grade compounds where possible, and university policies define one as a compound "for which a chemical purity standard has been established by a recognized national or regional pharmacopeia," like the USP (University of Arizona IACUC, 2023).

So even the strictest use of the phrase is about meeting a published standard. A seller writing "pharmaceutical grade" on a vial that no pharmacopeia standard covers, made in a plant nobody inspected, is borrowing a word, not meeting a rule.

"Research grade" has even less behind it. It isn't defined anywhere I looked. It usually just means "sold with a research-use disclaimer."

A lot, and most of it happens before and after the chemistry.

An approved drug has been through FDA review, and every batch is made under current good manufacturing practice, the rules in 21 CFR parts 210 and 211. Two of those rules matter most here:

Pyrogens are things that cause fever when injected. Bacterial endotoxin, a piece of dead bacteria's cell wall, is the main one. A vial can have zero living bacteria and still carry endotoxin. Keep that in mind for later.

Both are legal ways to make a drug that isn't the approved product. Neither is FDA-approved, and FDA says plainly it "does not verify the safety, effectiveness or quality of compounded drugs before they are marketed" (FDA).

503A is your local compounding pharmacy. The drug has to be made for "an identified individual patient based on the receipt of a valid prescription order" (21 USC 353a). These pharmacies are exempt from CGMP. For sterile products they follow USP chapter 797, the sterile compounding standard, whose current version became official November 1, 2023 (Pharmaceutical Technology, 2022).

Under 797, sterility and endotoxin testing depend on the preparation. Sterility testing is required for longer-dated preparations, and endotoxin testing for those made from nonsterile ingredients and for multiple-dose vials (ARL Bio Pharma, 2024). A peptide made from powder into a multi-dose vial usually lands in the second group.

503A also limits what raw powder a pharmacy can use. A bulk drug substance has to meet a USP or NF monograph, be a component of an FDA-approved drug, or be on FDA's 503A bulks list. It also has to come from an FDA-registered manufacturer and be "accompanied by valid certificates of analysis" (21 USC 353a). That's why it matters which peptides are on that list. The compounding status tracker follows where each one stands.

503B is an outsourcing facility that can make larger batches. It is subject to CGMP, inspected by FDA on a risk-based schedule (FDA), and has to file adverse event reports (21 USC 353b).

Less than vendors imply, and it's mostly not about drugs.

The famous line "For Research Use Only. Not for use in diagnostic procedures." comes from 21 CFR 809.10(c). That rule covers in vitro diagnostic products: test kits and reagents in the lab research phase (21 CFR 809.10). FDA's guidance on RUO labeling is a medical device guidance for diagnostics (FDA, 2013).

The drug-side version is narrower. A new drug "intended solely for tests in vitro or in laboratory research animals" can ship without an IND (21 CFR 312.2) if it carries this label: "CAUTION: Contains a new drug for investigational use only in laboratory research animals, or for tests in vitro. Not for use in humans." The shipper also has to keep records of "the name and post office address of the expert to whom the drug is shipped" (21 CFR 312.160).

Neither rule requires any quality testing. They're shipping exemptions, not quality grades.

And the label doesn't decide the question anyway. FDA looks at "objective intent," which can be shown by "labeling claims, advertising matter, or oral or written statements" (21 CFR 201.128). In a 2025 warning letter, FDA noted a peptide site's "research use only" and "not for human consumption" labels, then concluded the evidence "establishes that certain products ... are drugs intended for human use" (FDA warning letter, Feb 26, 2025). That's the same logic behind the five letters in September. More on the label itself in what RUO actually means.

A typical research-seller COA answers two questions: is it the right molecule (mass spec), and how pure is it (HPLC). That's covered in how to read a peptide COA.

What it usually doesn't answer is whether the vial is sterile or how much endotoxin is in it. HPLC purity measures what share of the peptide-like material is the target peptide. It tells you nothing about bacteria or endotoxin. Those take separate tests, and nothing requires a research seller to run them.

When those tests do get run, they find things. Researchers at the University of Pécs and UC San Diego bought semaglutide from online sellers without a prescription. Three of six orders arrived. All three vials were sterile when tested, but "endotoxin was detected in all samples," between 2.16 and 8.95 EU/mg. Purity measured 7.7% to 14.37%, against the 99% printed on the labels, and the semaglutide content ran 28.56% to 38.69% above the labeled amount (Ashraf et al., J Med Internet Res, 2024).

Read that again. No living bacteria, more semaglutide than the label claimed, and still endotoxin in every vial. A COA that only covers purity and identity would have missed it.

The bigger dataset I've written about, one lab's 6,441 samples, found a measurable endotoxin result in about 15% of the 243 samples where anyone paid for that test (BHN on the preprint). It hasn't been peer reviewed. The useful number there is the 243: only a small share of samples were ever checked for endotoxin.

Not automatically. The approved route has the most required testing. Compounded is in between, and the published testing on it is mostly done by the drug makers who lose sales to it, so weigh it accordingly.

Novo Nordisk scientists tested 16 injectable semaglutide follow-on products and found "new impurities and impurity patterns, including high molecular weight proteins, trace metals, anions, counterions, and residual solvents" compared with the original (Hach et al., Pharm Res, 2024). Lilly scientists reported "a widespread and previously unidentified impurity" in compounded tirzepatide mixed with vitamin B12, with clinical effects "unknown" (Jordan et al., Expert Opin Drug Saf, 2026).

FDA's own GLP-1 page, updated October 1, says it has warned companies selling semaglutide, tirzepatide, retatrutide and others "falsely labeled 'for research purposes' or 'not for human consumption,'" and that it "is aware of counterfeit Ozempic marketed in the U.S." (FDA, 2026).

Nobody can tell you, because nobody is required to find out. That's the honest answer, not a dodge.

A research vial has no approval review, no required batch testing and no required sterility or endotoxin check. In FDA's words, drugs outside the system "may be contaminated, counterfeit, contain varying amounts of active ingredients, or contain different ingredients altogether" (FDA warning letter, 2025). A 2026 medical review notes that BPC-157, TB-500 and retatrutide "lack established human safety profiles" (Moiz et al., Mo Med, 2026).

Case reports are starting to show up too. This month a Swiss team described a woman in her mid-20s hospitalized with days of vomiting after four weekly injections of "an unverified product reported to contain retatrutide," which doctors rated a probable cause (Buzzi and Huber, BMJ Case Rep, 2026).

Lyophilized means freeze-dried. FDA describes it as "a process in which water is removed from a product after it is frozen" (FDA inspection guide). It's why research peptides arrive as a powder puck.

It says nothing about grade. Approved drugs are lyophilized too. Egrifta WR, an approved tesamorelin, ships as a freeze-dried powder mixed with bacteriostatic water (FDA label, 2025).

What makes a lyophilized drug sterile isn't the freeze-drying. In FDA's description, the solution is first passed through "a 0.22 micron bacteria-retentive filter" and filled "under aseptic conditions." The same guide notes that inspections of these plants "have disclosed potency, sterility and stability problems" even under FDA oversight. A powder looking clean and dry tells you nothing about how it was filled. How long it lasts once mixed is its own question, covered in how long mixed peptides last.

Next step: see the vendors I use and how they score on testing, reviews and support. Code BHN40 takes 40% off.

If you've ever paid for an endotoxin or sterility test on a vial, reply and tell me what came back and what it cost. That's the data almost nobody has.