Sterility testing answers one narrow question: are viable microorganisms present in the sample being tested? It is one of the most misread lines on any certificate, because a "sterile" result is routinely treated as a blanket guarantee of safety, when the test was never designed to prove anything of the kind. This article explains what the USP <71> sterility test actually measures, how its two methods work, why "sterile" and "endotoxin-free" are separate properties, and why research-grade material is rarely sterility tested at all.
What sterility testing proves, and what it does not
The scope of the test is deliberately narrow. Under USP <71>, a product complies with the sterility test if there is no evidence of microbial growth after the incubation period. The method detects only viable microorganisms; it says nothing about chemical impurities or pyrogenic contaminants. In other words, a passing sterility result is evidence that no living bacteria or fungi grew from the containers that were tested. It is not a statement about identity, purity, potency, residual solvents, heavy metals, or endotoxin load.
That distinction is the single most important thing to carry into any certificate. Sterility is a microbiological property, and it is measured by growth, not by chemistry. A vial can be perfectly sterile and still be the wrong molecule, under-potent, or carrying a pyrogen that the sterility test is structurally incapable of seeing.
Inside USP <71>: the two compendial methods
The published USP <71> procedure recognizes two methods for detecting viable organisms: membrane filtration and direct inoculation (also called direct transfer). Membrane filtration is the preferred method whenever the product can be passed through a filter; direct inoculation is used for sample types that cannot be filtered.
Membrane filtration
In the membrane filtration method, the test article is passed through a sterile membrane filter with a pore size of 0.45 micron or less. Microorganisms, being larger than the pores, are retained on the membrane while the product passes through. The membrane is then transferred into culture media and incubated. Filtration is the preferred approach in part because it can help separate any organisms from components of the product that might otherwise inhibit their growth.
Direct inoculation
When a sample cannot be filtered, USP <71> uses direct inoculation, in which a measured amount of the test article is transferred directly into the culture media. The organisms, if present, then grow in the media itself. This is the fallback for materials whose form or chemistry makes filtration impractical.
The media and the 14-day incubation
Both methods rely on two culture media. Fluid Thioglycollate Medium (FTM) is used primarily for anaerobic bacteria and is incubated at 30 to 35 degrees C. Soybean-Casein Digest Medium, also called Tryptic Soy Broth (TSB), supports aerobic bacteria and fungi and is incubated at 20 to 25 degrees C. Using both media across both temperature ranges is what lets a single test cover the main classes of organism that could be present.
The incubation is not quick. USP <71> requires the test containers to be incubated for at least 14 days, specifically so that slow-growing microorganisms have time to become detectable. A result read too early could miss exactly the organisms the long incubation is designed to catch. At the end of that period, the compliance rule is simple: no evidence of growth means the sample passes.
Sterile is not the same as endotoxin-free
Here is where sterility is most often over-read. Passing a sterility test does not guarantee that a product is endotoxin-free. Sterility (USP <71>) and endotoxin content (USP <85>) are separate tests, and they have to be, because a sterilization process can kill bacteria without deactivating the endotoxins those bacteria leave behind.
The reason lies in what endotoxin is. Endotoxins are lipopolysaccharides that form the outer membrane of gram-negative bacteria, and the lipid A domain is the toxic, endotoxic portion of the molecule. This material is released when bacterial cells die or lyse, whether through autolysis, complement activation, phagocytosis, or the action of antibiotics. So the very act of killing bacteria, which is what a sterilization step does, can free endotoxin into the product. Endotoxin is a potent pyrogen: it triggers an immediate immune response and can cause fever, and at high exposure, septic (endotoxic) shock.
Endotoxin is also chemically stubborn in a way that live bacteria are not. It is heat-stable, and moist-heat autoclave sterilization does not destroy it; autoclaving can even liberate additional endotoxin by lysing bacterial cells. Removing it requires far harsher conditions than killing organisms does. Depyrogenation by dry heat at 250 degrees C for 30 minutes achieves at least a 3-log (1000-fold) reduction of endotoxin, a far more severe treatment than routine sterilization. The practical takeaway is that "sterile" and "low endotoxin" are earned by different processes and proven by different tests.
How endotoxin is measured and limited
Endotoxin has its own reference method. The USP <85> Bacterial Endotoxins Test uses Limulus Amebocyte Lysate (LAL), derived from horseshoe crab blood, and recognizes three method types: gel-clot, turbidimetric, and chromogenic. Unlike the sterility test, it is a biochemical assay for the presence and amount of endotoxin, not a test for living organisms.
The acceptable amount is not arbitrary. For parenteral drugs, the endotoxin limit is calculated as K divided by M, where K is the threshold pyrogenic dose (5 EU/kg for most routes of administration, and 0.2 EU/kg for intrathecal administration) and M is the maximum recommended human dose per kilogram of body weight per hour. That formula is why endotoxin limits are route-dependent and dose-dependent, and why a purity percentage tells you nothing about whether a material would meet them.
Why research-grade material is rarely sterility tested
Put these facts together and the gap in most research-grade documentation becomes obvious. Research-grade peptides are typically supplied as non-sterile lyophilized powder and are usually tested only for purity and identity, for example by HPLC at a purity of typically 95 percent or higher. Pharmaceutical-grade material is held to a longer list: in addition to identity and purity, it requires impurity profiling, endotoxin testing, sterility testing, and stability studies. Sterility and endotoxin are precisely the attributes that the shorter research-grade panel usually omits, which is why a purity COA does not cover them.
There is a regulatory logic to this as well. "Research use only" (RUO) products are labeled as not intended for human use, and they are not subject to the same FDA regulatory oversight or pharmaceutical manufacturing standards as prescription drugs. Sterility and endotoxin testing are lot-release expectations built for injectable medicines, not for powders sold for laboratory research, so their absence on a research-grade certificate is the norm rather than an anomaly. The regulatory backdrop for these health-product standards is explored further at the Coalition for Better Health.
None of this makes a purity-only certificate dishonest. It simply means the document has a defined scope. If sterility or endotoxin status matters for a given use, those are separate questions, answered by separate methods, and a laboratory has to be asked to run them. Choosing a lab therefore starts with knowing which tests a competent laboratory can actually perform, and reading any result inside the broader map of what safety testing does and does not cover.
The bottom line
Sterility testing under USP <71> proves one thing well: that no viable microorganisms grew from the tested containers over at least 14 days of incubation in two media, using membrane filtration where possible and direct inoculation where not. It does not prove chemical purity, it does not prove potency, and above all it does not prove the absence of endotoxin, which is heat-stable, released precisely when bacteria die, and governed by its own USP <85> limit. Research-grade material is rarely sterility tested because it is sold as non-sterile, research-use-only powder in the first place. Read a "sterile" claim for exactly what it certifies, and no more.