Microbial testing for supplements: what a Certificate of Analysis should show

Two kinds of microbial result, not one

A supplement COA's microbial section is not a single pass/fail line. It typically reports two structurally different kinds of test, and a reviewer checks them differently:

USP describes <2021> as providing "tests for the estimation of the number of viable aerobic microorganisms present in nutritional supplements of all kinds, from raw materials to the finished forms": an enumeration chapter, producing a count. <2022> is a different kind of test: it defines "objectionable" microorganisms as those that "represent a potential health hazard to the user" or are "capable of growing in the product," and sets out qualitative procedures to determine whether those specific organisms are present or absent, not how many.

The broader standard behind both: USP <2023>

USP General Chapter <2023>, Microbiological Attributes of Nonsterile Nutritional and Dietary Supplements, is the chapter that explains why any of this testing exists for products that are never intended to be sterile. Its own text notes that ingredients "may range from chemically synthesized vitamins to plant extracts and animal byproducts," and that botanicals in particular "may be microbiologically contaminated at any point during cultivation, harvesting, processing, packing, and distribution", citing sources such as "human or animal feces used as plant manure," contaminated irrigation or process water, and "poor worker hygiene and sanitation practices." The chapter's answer is not sterility but control: Good Manufacturing Practices, bioburden control on raw materials, and "adequate microbiological specifications" set for the finished product. <2021> and <2022> are how a manufacturer demonstrates that control is working on a given batch.

Reading the wording, not just the pass column

The same discipline that applies to reading any COA line applies here: the number or finding matters, not the lab's summary judgment of it. Two habits are worth checking on a microbial results table specifically:

• Enumeration results need a unit and a sample basis. A total aerobic count is only meaningful as, for example, CFU per gram. A bare number with no unit, or a result reported only as "pass," doesn't tell a reviewer what was actually measured.
• Specified-organism results need the sample size tested. “Absent” on its own is a weaker statement than “absent in [sample size]”: the sample size the absence was demonstrated against changes how much that finding actually rules out. This is the same logic as a detection limit on a chemical result: absence is only informative relative to how hard the method looked.
• “Conforms” is not a result for either kind. It tells you the lab was satisfied, not what it found. As with heavy-metal or potency results, a reviewer verifying a COA against an external requirement needs the actual count or the actual present/absent finding, not the lab's own summary of it, the same point covered in how to read a Certificate of Analysis.

Where the specific limits live

USP's general chapters establish the test methods and the categories of organism checked; the actual acceptance numbers vary by product type and are set out, product type by product type, with their own citations in the requirements library, USP. Neither this guide nor a lab's own marketing page is the place to look up a specific limit. The library row carries the citation and its verification date, which a number restated in an article does not.

Related: how a Certificate of Analysis is structured overall, including non-detect notation on the chemical side, is in how to read a Certificate of Analysis; how heavy-metal results are checked is in heavy metals testing for supplements. Compliant Always's free checkreads both the count and the specified-organism lines off an uploaded COA and flags results reported only as “conforms.”