Endotoxin vs sterility: two tests that answer different questions

Endotoxin vs sterility at a glance

An endotoxin test asks how much of one bacterial molecule is in the sample. A sterility test asks whether anything in the sample is alive. The two use different methods, different samples and different time scales. Neither result stands in for the other.

	Endotoxin	Sterility
Question	How much bacterial endotoxin is in the sample?	Do live bacteria or fungi grow from the tested units?
What it detects	Lipopolysaccharide from Gram-negative bacteria, alive or dead	Living aerobic and anaerobic bacteria, yeasts and moulds
Standard method	LAL test, USP <85>	Sterility test, USP <71>
How it works	Horseshoe crab lysate reacts with endotoxin; the reaction is read against a standard	The sample goes into two culture media, which are watched for growth
Time	About an hour for the gel-clot reaction	At least 14 days of incubation
Result	A number: EU per mg and EU per vial	Growth or no growth in the units tested
Sample	A measured portion of the dissolved sample	Whole unopened units, in a number set by the method
What it cannot show	Whether live microbes are present	How much endotoxin is present, or whether untested vials are sterile

Both chapters are harmonised with the European and Japanese pharmacopoeias, so the same methods apply in all three, apart from a few marked local points.

Why a vial can pass one test and fail the other

Sterile, with endotoxin

Endotoxin sits in the outer membrane of Gram-negative bacteria. An FDA inspection guide notes that bacteria release it when the cell dies and breaks apart. The same guide says endotoxin is not removed by sterilising or microbiological filters, so lowering the number of microbes does not lower the endotoxin with it. A product can therefore contain no living cells and still contain endotoxin from bacteria that grew earlier in production.

Clearing endotoxin needs its own step. The pharmacopoeia’s endotoxin chapter has labs depyrogenate glassware with dry heat, commonly at least 30 minutes at 250 °C, and the FDA guide cites reports of 180 °C for 4 hours. A sterility pass cannot see endotoxin at all.

Low endotoxin, not sterile

The harmonised chapter defines the LAL test as a test for endotoxins of Gram-negative bacterial origin. Fungi and Gram-positive bacteria do not make endotoxin. The sterility method checks its media against six test organisms, including the yeast Candida albicans, the mould Aspergillus brasiliensis and the Gram-positive Staphylococcus aureus. A vial with a low endotoxin number can still hold live microbes of these kinds.

Neither covers every fever-causing substance

Endotoxin is one kind of pyrogen, a substance that causes fever. FDA’s 2026 questions and answers say makers of drug materials and devices should assess the risk of non-endotoxin pyrogens, and that the rabbit pyrogen test may suit better where those may be present. An LAL result does not cover them.

How each test works

The endotoxin test: LAL, USP <85>

The lab mixes a dilution of the sample with Limulus amebocyte lysate, an extract of horseshoe crab blood cells. In the gel-clot method, the tubes incubate at 37 °C for 60 minutes and are then inverted: a firm gel that stays in place is positive. Photometric methods read cloudiness or colour against a standard curve and give a number. In those methods the run counts only when endotoxin added to the sample is recovered at 50% to 200%, which shows the peptide did not hide or inflate the reading. Our endotoxin testing guide covers units, limits and the spike recovery in detail.

The sterility test: USP <71>

The harmonised method sets out each step:

• Two media. Fluid thioglycollate medium, incubated at 30 to 35 °C, is mainly for anaerobic bacteria and also detects aerobes. Soya-bean casein digest medium, incubated at 20 to 25 °C, is for fungi and aerobic bacteria.
• Two techniques. Membrane filtration is used whenever the product allows. The dissolved sample passes through a filter with a nominal pore size of 0.45 µm or less, and the filter goes into the media. Otherwise the sample goes straight into the media (direct inoculation).
• At least 14 days. The media incubate for not less than 14 days and are examined at intervals. If the sample clouds the medium so that growth cannot be judged by eye, portions move to fresh medium at day 14 and incubate for at least 4 more days.
• A suitability check. For each new product, and whenever conditions change, the lab runs the method with no more than 100 colony-forming units of each test organism added. Growth comparable to a control without product shows the product does not stop microbes from growing. This is the sterility counterpart of the endotoxin spike recovery.
• A clear decision rule. No growth means the product complies. Growth means it does not comply, unless the test is shown to be invalid for one of four listed reasons, such as growth in the negative controls. An invalid test is repeated with the same number of units.

What a sterility pass covers

The harmonised text states the limit plainly: a satisfactory result only shows that no contaminating organism was found in the sample examined, under the test conditions.

For injectable products, the method sets a minimum number of units to test from each batch:

Units in the batch	Minimum units tested
Up to 100	10% or 4, whichever is greater
101 to 500	10
More than 500	2% or 20, whichever is less

If the batch size is not known, the maximum applies. The count is per medium, unless one container holds enough for both media. For a solid of less than 50 mg per container, each medium gets the whole contents of a container, so one such vial cannot supply both. A 10 mg vial falls in that row. A buyer who sends two vials gets a result on two vials. Our guide on how many vials to send shows what each lab asks for.

When a lab’s “sterility” is a different test

In our September 2026 review of public certificates from 25 labs and one rating site, none showed a documented 14-day USP <71> test with the number of units tested, and one showed it in part. Tests printed as “sterility” included microbial counts, PCR screens, 2-day cultures and a citation of a method-validation chapter. Each of these answers a real question. None is the USP <71> test.

• Microbial counts (TAMC and TYMC). These belong to the harmonised tests for non-sterile products (USP <61>). They count bacteria and fungi that grow under aerobic conditions and compare the count with a limit. They do not look for anaerobes. Our guide to bioburden vs sterility explains counts in more detail.
• PCR or DNA screens. These look for microbial DNA in place of growth, and they answer sooner.
• Short cultures. A 2-day reading uses growth, but it stops 12 days before the method’s minimum. It is a preliminary result.

These are the names labs used on their own sites in September 2026:

Lab	Name on the lab’s site	What it is
Krause Analytical	“Sterility, 14 Day - USP 71”, an add-on	USP <71>, 14 days
Vanguard Laboratory	Sterility (USP 71, 14-day), an option on every package	USP <71>, 14 days; bioburden (USP 61) is a separate test in its Gold package
Accumark Labs	“Sterility USP” and “Rapid Sterility Screening (PCR)”	USP <71> with a dedicated vial and a 14-day turnaround; a PCR screen sold as a separate add-on
Janoshik	“Sterility testing (TAMC+TYMC)”	A microbial count by culture, reported as pass or fail
Freedom Diagnostics	“Sterility (PCR)”	A PCR screen for microbial contamination, as the lab describes it

Labs that sell a screen next to USP <71> and name each one help the buyer choose. On any report, check the method name, the number of days and the number of units before you read the result. Our guide to sterility testing for peptides goes further.

What to look for on the report

Item	Endotoxin	Sterility
Method	USP <85> and the LAL technique: gel-clot, turbidimetric or chromogenic	USP <71>, by membrane filtration or direct inoculation
Result	EU per mg and EU per vial, or “less than” a stated limit of quantitation	Growth or no growth in each medium
Validity check	Spike recovery between 50% and 200%	Suitability check with test organisms, and negative controls
Scale	Replicate readings	Number of units tested
Dates	Date of the test	Start and end of incubation, at least 14 days apart
Scope	Which vial was used	A statement that the result covers only the units tested

Our page on how to read a COA covers the rest of the report.

What HK Analytical offers for each

Our endotoxin test uses LAL and reports EU per mg and EU per vial, with the limit of quantitation, replicates and spike recovery. Our Sterility (USP <71>, 14 days) test incubates the tested units for at least 14 days, and the result covers only those units. Sterility needs unopened vials of its own, and your quote states how many. Order either test or both, in any mix with the rest of the panel, per sample. Results within days of receipt. Sterility (USP <71>, 14 days) is the one test that takes 14 days, because the method incubates for 14 days. Every number comes with its raw data, in a signed record you can check on /verify. See the example report or request a quote.