The short version: endotoxin testing [USP <85>, the LAL assay] measures a heat-stable bacterial residue that survives filtration. PCR screening looks for microbial DNA above a detection floor.
Sterility testing [USP <71>] is a roughly 14-day culture that looks for living organisms and destroys the vials it tests — and almost nobody in this market runs it.
The three are not interchangeable, a pass on one says nothing about the others, and for injectable-format compounds the minimum a buyer should demand is lot-specific endotoxin data reported as a number against a stated threshold.
- Endotoxins are lipopolysaccharide [LPS] fragments of dead gram-negative bacteria — LPS being the sugar-and-lipid material of the outer bacterial wall. They are heat-stable, fever-inducing, and able to pass straight through the sterile filter that removed the bacteria themselves.
- The LAL assay [USP <85>] measures that residue against a threshold. Duplicates and controls are what make the number trustworthy — a single uncontrolled reading is close to decoration.
- A PCR screen detects microbial DNA. It can flag dead organisms and miss live ones below its detection floor — it is a contamination screen, not a sterility result.
- USP <71> sterility testing takes about 14 days and destroys the units it tests. That is why the research market almost never runs it — and why casual "sterility assured" language should be read as marketing.
- "Sterile" on bacteriostatic water is a form descriptor about a manufacturing process. It is not a test result on your peptide lot, and vendors trade on the confusion.
The layer under the certificate
This report explains the microbiology underneath the certificate. The COA guide explains the certificate itself.
Our guide to reading a certificate of analysis covers the five measures a trustworthy document reports and how to verify one.
This piece assumes that and goes one level down, into the microbiology the certificate is summarizing — because the market's favorite trick is not fake documents.
It is real documents describing modest tests, sold under words that imply bigger ones. To see through that, you need to know what each test physically does.
Endotoxin survives sterilization
Endotoxins are cell-wall fragments shed by dead gram-negative bacteria. They cause fever, survive heat, and pass straight through a sterilising filter.
Start with the molecule, because everything else follows from its properties. Endotoxins are lipopolysaccharide fragments — pieces of the outer cell wall of gram-negative bacteria. They are shed while the organisms are alive and released in bulk when they die. Three properties make them the central quality question for anything in injectable format:
- They are pyrogenic at tiny doses. The mammalian immune system treats lipopolysaccharide as a primary invasion signal; parenteral exposure [exposure by injection, bypassing the gut] drives fever and inflammatory response at doses measured in fractions of a microgram.
- They are heat-stable. Conditions that kill every living organism in a vial leave endotoxin largely intact. You cannot autoclave your way out of an endotoxin problem.
- They pass sterile filters. A 0.22-micron filter removes bacteria. The wall fragments of bacteria that died upstream — in the water, in the synthesis, on the equipment — sail through.
Hold those three facts together and the market's core confusion dissolves: a lot can contain zero living organisms and still be pyrogenic. Killing the bacteria is not the same as removing what their walls left behind.
This is why endotoxin gets its own pharmacopeial chapter, its own assay, and its own line on serious certificates. It is also why no sterility claim, true or false, answers the endotoxin question.
The LAL assay
USP <85> measures endotoxin using a horseshoe-crab reagent. Only a number, a stated threshold and duplicate readings make the result worth anything.
The standard method is the bacterial endotoxins test of USP <85>, built on LAL. That is limulus amebocyte lysate, a reagent derived from horseshoe crab blood cells, which clots in the presence of endotoxin with remarkable sensitivity.
The assay comes in gel-clot and photometric variants, and results are reported in endotoxin units against a threshold set by the product's dose and route.
When you see LAL data on a certificate, the reporting format tells you how seriously the lab took it. Three things separate a real result from a checkbox:
- A number, not a word. "<0.5 EU/mg" [endotoxin units per milligram] is a measurement. "Endotoxin: pass" is a claim about a measurement you are not being shown.
- A stated threshold. A number only means something against the limit the lab applied — a serious document states both.
- Duplicates and controls. Good labs run samples in duplicate with positive product controls, because peptide matrices can inhibit or enhance the reaction and quietly corrupt a single reading. Duplicate agreement is what converts a reading into a result.
One development worth knowing about: recombinant factor C [rFC] reagents — synthetic alternatives to crab-derived lysate — have moved from the sustainability literature into the pharmacopeias. USP has recognized recombinant-reagent methods in a dedicated chapter.
For a buyer the distinction is minor. An endotoxin number from a competent lab is the point, whichever reagent produced it.
What PCR cannot see
A PCR screen detects microbial DNA. It can miss living organisms below its detection floor and flag organisms that died during processing.
The second test that shows up on better certificates is a PCR microbial screen — amplification chemistry looking for microbial DNA in the sample. It is a real test with real value, and it is routinely oversold, so be precise about both directions of its blindness.
What it can miss: PCR has a detection floor. Viable organisms present below that floor — or organisms whose signatures the panel does not target — produce a clean result. "No detectable microbial DNA" is a statement about the assay's sensitivity, not a guarantee of absence.
What it can falsely flag: DNA does not die when its organism does. A sample can carry amplifiable DNA from organisms killed long ago in processing — harmless residue reading as a hit. PCR cannot distinguish a live contaminant from a ghost.
Put both limits together and the correct reading is: a PCR screen is a useful contamination check that is neither necessary nor sufficient for sterility. On a certificate it earns respect as an extra layer, never as a substitute for the two tests on either side of it.
What sterility testing costs
Sterility testing runs at least 14 days, destroys the vials it samples, and must itself be performed under cleanroom conditions. That is why this market skips it.
True sterility testing is USP <71>, and its mechanics explain the market's behavior better than any accusation could. The method incubates product in two growth media — one favoring bacteria, one favoring fungi — for roughly 14 days, watching for any growth. Three consequences follow:
- It is slow. Two weeks of incubation before a lot can claim the result — an eternity for vendors whose inventory turns weekly.
- It is destructive. The vials tested are consumed. Sampling enough units to make the result meaningful means destroying sellable product from every lot.
- It punishes sloppiness in both directions. The test must itself be run under contamination-controlled conditions, because a technician's stray organism condemns a clean lot. Labs that run <71> properly are running cleanroom operations to test for cleanroom failures.
Pharmaceutical manufacturers absorb all three costs because regulation requires it. Research-market vendors face no such requirement, and almost none volunteer. That is rational economics, and worth saying plainly rather than darkly. The problem is not that vendors skip <71>. It is that their marketing implies they did not.
When a product page says "sterility assured" and the certificate shows an LAL number and a PCR screen, the words and the documents disagree. The documents are the ones telling the truth.
Sterile versus sterility
On a diluent, sterile describes a manufacturing process. On a peptide lot, sterility is a test result. Only one of the two says anything about your vial.
One more distinction the market blurs, this one subtler. Bacteriostatic water and other diluents sold alongside peptides are commonly labeled sterile — legitimately.
On those products, "sterile" is a form descriptor. It says the product is manufactured and filled under an established sterile process, a claim about how a production line operates. A sterility test result on a peptide lot is a different object entirely — a measurement performed on that specific lot after filling.
Both usages are legitimate in their own lanes. The trouble is adjacency. A vendor page showing sterile-labeled diluent next to peptide vials lets the adjective drift sideways onto products it was never tested on.
When you read "sterile" anywhere in this market, the question to ask is mechanical. Is this word describing a process claim on a diluent, or a <71> result on the lot in my hand. They are almost never the second.
Where aseptic fill fails
With no sterility testing, filtration and aseptic filling are the only defence. Every failure point in that chain is a human one.
Since almost no research lot is sterility-tested, what stands between a buyer and contamination is process: sterile filtration plus aseptic fill.
In practice that means passing the reconstituted bulk solution through a 0.22-micron filter, directly into vials, inside an environment intended to keep new organisms out. That filter removes organisms, not endotoxin. Capping follows, and typically lyophilization [freeze-drying back to a dry powder].
The chain fails at its human links. The filter itself is reliable. What fails is everything around it. Vials or stoppers go inadequately depyrogenated [not treated to destroy the endotoxin already sitting on them]. Filling environments are a bench and a wiped-down hood rather than a qualified cleanroom. Operators shed organisms over open vials. And endotoxin has already accumulated upstream in the water or the raw powder before filtration ever happened.
This is why the two cheap tests are so informative despite their limits. An endotoxin number audits the upstream water-and-handling chain, and a PCR screen samples the downstream fill environment.
Where your raw powder was synthesized — and how many hands it passed through before filling — is its own story, covered in our supply-chain report.
What to demand
Ask for lot-specific endotoxin data, as a number against a stated threshold, from a named lab. Treat anything short of that as untested.
For injectable-format research compounds, the reasonable floor is not sterility testing. Demanding <71> from this market is demanding a document nobody will produce.
The reasonable floor is lot-specific endotoxin data, reported as a number against a stated threshold, on a certificate naming the lab. That certificate is verifiable per the steps in our COA guide.
A PCR microbial screen on the same document is a meaningful upgrade. Anything beyond that is a bonus so rare it functions as a vendor-identification feature.
And any adverse reaction to a compound with no such data behind it belongs in the side-effects framework. Treat it as an unknown-quality event, not just an unknown-compound event. The vial, not the molecule, may be the variable.
The field guide
Five phrases the market uses interchangeably, and the exact claim each one does and does not commit to.
| test | method | what a pass means | what it cannot tell you |
|---|---|---|---|
| Endotoxin | LAL · USP <85> [or rFC] | Pyrogenic bacterial residue sits below the stated threshold in the tested sample. | Whether viable organisms are present. Not a sterility result. |
| Microbial screen | PCR amplification | No microbial DNA detected above the assay's floor, for the targets on the panel. | Live organisms below the floor; dead-DNA hits; anything off-panel. |
| Sterility | USP <71> · 14-day dual-media culture | No viable growth from the sampled units across the incubation. | Endotoxin load; anything about the unsampled vials in the lot. |
| "Sterile" diluent label | form descriptor | The diluent is produced under an established sterile-fill process. | Anything at all about the peptide lot sold next to it. |
| "Lab tested" [marketing] | unspecified | Some test, on some sample, at some point. Nothing more until the certificate says which. | Everything. The phrase commits to nothing. |
The last two rows are the market-language audit in miniature. "USP grade" deserves the same squint. USP grade describes a material meeting a monograph's standards. It is a purity-and-identity claim about an ingredient, not a microbiological result on a finished lot.
Every one of these phrases has a legitimate home. The audit question is always whether the phrase and the document underneath it are the same size.
FAQ
What is an endotoxin?
A heat-stable lipopolysaccharide fragment from the wall of a gram-negative bacterium. It is fever-inducing at tiny parenteral doses, and able to pass through the sterile filter that removed the bacterium itself. Those fragments are shed while the organisms are alive and released in bulk when they die.
Three properties make them the central quality question for anything in injectable format. They are pyrogenic at tiny doses. The mammalian immune system reads lipopolysaccharide as a primary invasion signal, and answers with fever and inflammation at fractions of a microgram.
They are heat-stable, so conditions that kill every living organism in a vial leave the endotoxin largely intact. You cannot autoclave your way out of the problem.
And they pass a 0.22-micron filter, which removes bacteria. It does not remove the wall fragments of bacteria that died upstream in the water, in the synthesis, or on the equipment. A lot can hold zero living organisms and still be pyrogenic.
Is an endotoxin test the same as a sterility test?
No. USP <85> measures a bacterial residue. USP <71> cultures for living organisms over roughly 14 days. A lot can pass either and fail the other, which is exactly why serious documents report them separately.
The endotoxin test is the LAL assay — limulus amebocyte lysate, a reagent derived from horseshoe crab blood cells that clots in the presence of endotoxin. It reports endotoxin units against a threshold set by the product's dose and route.
The sterility test incubates product in two growth media, one favoring bacteria and one favoring fungi, and watches for growth. It destroys the vials it samples.
Because endotoxin is heat-stable and passes sterile filters, a lot with no viable organisms can still carry a pyrogenic load. And a clean endotoxin number still says nothing about a live contaminant. Neither result substitutes for the other.
Why does almost nobody in this market run USP <71>?
It takes about two weeks, destroys the vials it samples, and requires contamination-controlled test conditions where a technician's error condemns a clean lot. Unregulated economics reward the fast cheap tests instead — so sterility language usually turns out to be endotoxin or PCR wearing a bigger word.
Take the three costs in turn. Two weeks of incubation is an eternity for vendors whose inventory turns weekly. Sampling enough units to make the result meaningful means destroying sellable product out of every lot.
And labs that run <71> properly are running cleanroom operations in order to test for cleanroom failures. Pharmaceutical manufacturers absorb all three costs because regulation requires it.
Research-market vendors face no such requirement and almost none volunteer, which is rational economics and worth saying plainly. The problem is not that vendors skip <71>. It is that their marketing implies they did not.
Does "sterile" on bacteriostatic water cover the peptide sold next to it?
No. On a diluent, sterile is a form descriptor about that product's manufacturing process. It transfers nothing to a peptide lot, which would need its own <71> result to make the equivalent claim.
The adjective says the diluent is manufactured and filled under an established sterile process — a claim about how a production line operates. A sterility test result on a peptide lot is a different object entirely: a measurement performed on that specific lot after filling.
Both usages are legitimate in their own lanes. The trouble is adjacency. A vendor page showing sterile-labeled diluent beside peptide vials lets the adjective drift sideways onto products it was never tested on.
So when you read sterile anywhere in this market, ask the mechanical question. Is this word describing a process claim on a diluent, or a <71> result on the lot in my hand. It is almost never the second.
What is the minimum testing worth insisting on?
Lot-specific endotoxin data — a number against a stated threshold, from a named lab, on a verifiable certificate. A PCR microbial screen is a strong addition. Treat sterility claims as unproven unless USP <71> appears on the document itself.
Demanding <71> from this market is demanding a document nobody will produce. So the reasonable floor sits one step below it: an LAL result, or a recombinant-reagent equivalent, run on your lot.
Three details separate a real result from a checkbox. A number rather than a word, because endotoxin: pass is a claim about a measurement you are not being shown.
A stated threshold, because a number only means something against the limit the lab applied. And duplicates with positive product controls, because peptide matrices can inhibit or enhance the reaction and quietly corrupt a single reading.
References
- USP General Chapter <85> Bacterial Endotoxins Test — usp.org
- USP General Chapter <71> Sterility Tests — usp.org
- USP, recognition of recombinant-reagent bacterial endotoxins methods [rFC] — usp.org
- FDA, Guidance for Industry — Pyrogen and Endotoxins Testing: Questions and Answers — fda.gov
- FDA, Guidance for Industry — Sterile Drug Products Produced by Aseptic Processing, Current Good Manufacturing Practice — fda.gov
disclosure: this is an evidence piece — it contains no vendor links. inside your peptides is published by the owners of the next lab, the only vendor we are paid by elsewhere on the site; grades and rankings follow the published criteria in our editorial policy — never referral terms. research + education only · not medical advice.