Endotoxin (LAL) testing explained: what an EU/mg result on a COA means

Q: Does a high HPLC purity mean a sample is low in endotoxin?

No. HPLC purity measures peptide-related impurities. Endotoxin is a bacterial contaminant that HPLC does not detect, so it needs its own test.

Q: What does a result of less than 0.5 EU/mg mean?

It means endotoxin was below the level the test could detect or quantify at the dilution used. It does not mean zero, and the figure depends on the method sensitivity and dilution.

Q: Is a sterile-filtered sample free of endotoxin?

Not necessarily. A 0.22 µm filter removes bacteria but is not designed to remove endotoxin, which is also heat-stable. Sterility and endotoxin are tested separately.

Q: Is LAL or recombinant Factor C better?

Both are recognised. LAL methods are in USP . Recombinant reagents are in USP , official since May 2025 as an alternative method. rFC avoids the glucan pathway in LAL, but either method must be shown to suit the sample.

Published 3 Nov 2026 · By The QLD Analytical Team

Informational guide for research samples — not medical advice. QLD Analytical arranges independent testing; we do not sell or endorse any product.

Most certificates of analysis for research peptides report identity and purity. Some also carry an endotoxin result, usually a number in EU/mg or a “less than” value. This guide explains what endotoxin is, how the main test methods work, and what to check before you rely on an endotoxin figure.

What endotoxin is

Bacterial endotoxins are lipopolysaccharides (LPS) from the outer membrane of Gram-negative bacteria. They are shed as bacteria grow and released in large amounts when the cells die and break apart. The part of the molecule responsible for most of its biological activity is lipid A.

For a synthetic peptide, endotoxin is a contamination question rather than a chemistry question. It can come from water, buffers, glassware, equipment or handling at any stage of synthesis, purification, freeze-drying and filling. Because endotoxin is not a peptide-related impurity, it does not show up in the tests most COAs report. HPLC purity, mass spectrometry identity and net content can all look excellent on a sample that carries significant endotoxin.

Why “sterile” does not mean endotoxin-free

Sterility and endotoxin are two separate properties:

Sterile means there are no living micro-organisms that can grow.
Endotoxin is a molecule left behind by bacteria, whether they are alive or dead.

Filtering a solution through a 0.22 µm membrane removes bacteria, but it is not designed to remove endotoxin. Endotoxin is also very heat-stable: ordinary autoclaving does not reliably destroy it, which is why glassware is depyrogenated with dry heat at much higher temperatures. A product can therefore be sterile and still fail an endotoxin test, or contain no bacteria at all and still carry the residue of earlier contamination.

How the LAL test works: gel-clot, turbidimetric and chromogenic

The standard bacterial endotoxins test (BET) uses Limulus amebocyte lysate (LAL), a reagent made from the blood cells of the Atlantic horseshoe crab. Endotoxin activates an enzyme cascade in the lysate that starts with Factor C and ends in clotting. The test methods described in USP chapter <85> read that reaction in three different ways:

Gel-clot. The sample is mixed with lysate of a known sensitivity and incubated. A firm clot that survives inverting the tube is positive. It is essentially a limit test: the result says whether endotoxin is above or below a threshold, rather than giving a precise number.
Turbidimetric. An instrument tracks the increase in cloudiness as the clot forms. In the kinetic version, the time taken to reach a set turbidity is compared with a standard curve to give a quantitative result.
Chromogenic. A synthetic substrate releases a yellow colour when it is cut by the activated enzyme. The colour is read photometrically, either at a fixed end point or kinetically, and again compared with a standard curve.

All three are accepted compendial methods. The kinetic methods give a number across a range, which is more useful when you want to compare samples or batches. The FDA’s guidance Pyrogen and Endotoxins Testing: Questions and Answers covers the gel-clot, photometric and kinetic approaches.

Recombinant Factor C and other animal-free methods

Newer reagents copy the first step of the horseshoe crab cascade using recombinant proteins. Recombinant Factor C (rFC) produces a fluorescent signal when endotoxin binds, and recombinant cascade reagents (rCR) rebuild more of the original pathway. USP chapter <86>, Bacterial Endotoxins Test Using Recombinant Reagents, became official in May 2025. USP treats it as an alternative to <85> that must be shown to suit the specific material being tested.

One practical difference is specificity. LAL also contains a second pathway (Factor G) that can respond to (1→3)-β-D-glucans, a possible source of false positives unless a glucan-blocking buffer is used. rFC does not have that pathway. When a COA names the method, it helps to know which reagent family was used.

Note that the BET detects endotoxin only. Other fever-causing substances (non-endotoxin pyrogens) need different tests, such as the monocyte activation test.

EU, EU/mg and the “less than” sign

Endotoxin is reported in endotoxin units (EU). EU measures biological activity relative to a reference standard endotoxin, not the mass of LPS. Results are normalised to the sample. For a powder or a vial contents this is usually EU/mg of material, and for a liquid it is EU/mL.

Many results are written as an inequality, such as “< 0.5 EU/mg”. That means the endotoxin was below the lowest level the test could detect or quantify at the dilution used. It does not mean zero. The figure after the “<” depends on the reagent’s sensitivity and on how far the sample had to be diluted, so two labs can report different “less than” values for equally clean material.

Interference, spike recovery and false results

Peptide samples can interfere with the LAL reaction. Extremes of pH, high salt, chelating agents or the peptide itself can suppress the signal (inhibition) or exaggerate it (enhancement). A suppressed signal is the more dangerous of the two, because it can make a contaminated sample look clean.

Laboratories control for this in two ways:

Dilution. Diluting the sample usually removes interference. There is a limit, the maximum valid dilution, beyond which the test can no longer detect endotoxin at the level that matters.
Positive product control (spike recovery). A known amount of endotoxin is added to a portion of the sample. If the test does not recover a sufficient fraction of that spike, the result is not valid. For the quantitative methods in USP <85>, recovery is expected to fall within 50–200%.

A full laboratory report usually lists the dilution used and the spike recovery. If a certificate gives only a bare “pass” with no method, dilution or recovery, it is hard to judge how much weight the result can carry.

What a “pass” is measured against

An endotoxin result is only a pass or a fail against a stated limit. For registered medicines, limits are set in pharmacopoeial monographs or calculated from the product’s intended route and amount of use. Research peptides have no single universal limit. For in-vitro work the acceptable level depends on the assay, and cell-based work with immune cells is often especially sensitive to LPS.

So when you read a COA, look for the limit or specification alongside the number. A result of “1.2 EU/mg” on its own is a measurement. “1.2 EU/mg, specification ≤ X EU/mg” is a judgement you can check.

Checklist for an endotoxin result on a COA

Is the method named? Look for gel-clot, kinetic turbidimetric, kinetic chromogenic or recombinant (rFC/rCR).
Are the units clear? Look for EU/mg or EU/mL, not just “pass”.
Is a limit stated, and does the result meet it?
For a “<” result, is the detection level shown? Check that it is low enough to be meaningful.
Is spike recovery or a valid positive product control reported, or at least stated to have been met?
Do the sample ID, lot and test date match the vial and the rest of the COA?
Is the laboratory named, and can the report be verified?

Our guides to what a COA should show and reading an HPLC chromatogram cover the other sections of a certificate.

Endotoxin testing through QLD Analytical

QLD Analytical is an independent testing intermediary. Samples are analysed offsite by Apex Labs Pty Ltd. Our standard panel covers HPLC purity, mass spectrometry identity and net content. Endotoxin (LAL) testing is arranged on enquiry, so please email orders@qldanalytical.com before you post. It needs a second vial of the same sample. Packing and postage details are in our customer guide and FAQ.

As with every test we arrange, an endotoxin result applies only to the sample sent in. It is not a guarantee about any other vial or batch, or about how a vial is stored or handled after testing.

FAQ

Does a high HPLC purity mean a sample is low in endotoxin?