Guide · Microbiology
Microbial limits on a CoA: cfu/g, absent in 25 g and more
- Short answer
- Microbial results are either counts or presence tests. A count such as 150 cfu/g is the number of colony-forming units per gram under the stated method, and a limit such as NMT 1,000 cfu/g caps it. Absent in 25 g means the organism was not detected in a 25 g test portion. Check the unit, the portion and the method.
General background, not regulatory or GMP advice. Acceptance criteria, test methods and sampling plans are set by your specification and the standard it cites.
Counts, absence and the conventions around them
Two kinds of result appear. Counts, in colony-forming units per gram (cfu/g) or per millilitre (cfu/mL), come from plate or filtration methods: a total aerobic count, a yeast and mould count, sometimes coliforms. Presence tests say an organism such as Salmonella, E. coli or Staphylococcus aureus was absent from a stated portion. Pharmaceutical certificates usually write the counts as TAMC (total aerobic microbial count) and TYMC (total combined yeasts and moulds count), terms used in USP <61> and the matching European Pharmacopoeia chapter 2.6.12.
Pharmacopoeial limits use a power-of-ten convention that surprises many readers. In USP <1111>, acceptance criteria written as 10¹, 10² or 10³ cfu are read as maximum acceptable counts of 20, 200 and 2,000. A limit of 10³ cfu/g can therefore be met by a result of 1,500 cfu/g, while a limit written NMT 1,000 cfu/g cannot. Know which convention your specification uses, and see reading limit symbols.
Reading a microbial section
Seven checks, line by line.
Sort the lines
Decide for each line whether it is a count (cfu/g), a presence test (absent in 25 g) or something else. Endotoxin in EU/mg, for example, is a separate test and not a colony count.
Read the limit's wording
NMT 1,000 cfu/g and ≤ 1,000 include 1,000. < 10 cfu/g excludes 10. A bare 10³ cfu/g may carry the pharmacopoeial meaning described above.
Check unit and portion
cfu/g and cfu/mL are not interchangeable, and absent in 10 g is weaker assurance than absent in 25 g. The result's portion should be at least as large as the specification's.
Read the form of the result
A number, '< 10 cfu/g', Absent or Not detected. A result of '< 10' is a reporting level, not zero. TNTC, too numerous to count, is not a result.
Check the method
Look for a reference such as USP <61> and <62>, Ph. Eur. 2.6.12 and 2.6.13, or for food ISO 4833-1 (aerobic count) and ISO 6579-1 (Salmonella). A different method can give a different count. See pharmacopoeial methods.
Check the sampling
A result describes the portion tested. Some food specifications state a sampling plan such as n = 5, c = 0: five sample units tested, none allowed to be positive. Ask what the certificate's result represents.
Compare with your own specification
Use your list of organisms, limits and portions, not just the supplier's. How buyers follow up is covered in trust but verify.
How microbial limits are written
Common forms, how they read, and what to watch.
| Written as | Reads as | Watch for |
|---|---|---|
| TAMC NMT 1,000 cfu/g | Total aerobic count must not exceed 1,000 per gram | The limit includes 1,000; check the method |
| TYMC NMT 100 cfu/g | Combined yeast and mould count must not exceed 100 | A separate count with its own limit |
| 10³ cfu/g | In pharmacopoeial usage, a maximum acceptable count of 2,000 | Not the same as NMT 1,000 |
| < 10 cfu/g (result) | None counted at the lowest level the method reports | A reporting level, not proof of zero |
| Absent in 25 g | Not detected in a 25 g test portion | The portion must match; a smaller one is weaker |
| n = 5, c = 0, absent in 25 g | Five units tested, none may be positive | Ask how many units the certificate covers |
| TNTC | Too numerous to count at that dilution | Not a result; ask for a countable dilution |
What goes wrong with microbial results
- A power of ten read as a plain number. 10³ cfu/g and NMT 1,000 cfu/g are not interchangeable.
- A smaller portion accepted. Absent in 1 g does not show absence in 25 g.
- cfu/g set against cfu/mL. For a liquid, converting needs a stated density, so ask the supplier.
- < 10 read as zero. It means none were counted at the lowest level the method reports.
- TNTC accepted as high but fine. It is not a result.
- One result read as lot assurance. Contamination can be uneven, so one tested portion cannot guarantee the rest of the lot.
How CertiClean helps
CertiClean reads counts and presence lines like any other row. A '< 10 cfu/g' result passes a maximum when the printed value is within it, Absent or Not detected meets an absence limit, TNTC is marked could not evaluate, and cfu/g against cfu/mL is flagged as a unit mismatch. It does not apply the power-of-ten convention (NMT 10³ cfu/g is read as 1,000, and a bare 10³ cfu/g limit is marked could not evaluate) and does not compare test portions, so those two checks stay with you. See CoA to Excel and pricing.
Free allowance: 1 file per day, up to 10 pages, no account.
Questions people ask
What does cfu/g mean?
Colony-forming units per gram: the number of viable microorganisms in a gram of sample that formed visible colonies under the stated method, medium, temperature and time. A different method can give a different count.
What does absent in 25 g mean?
The organism was not detected in a 25 g test portion. It is a presence or absence result, not a count, and it applies to the portion tested. A smaller portion gives less assurance.
What does < 10 cfu/g mean?
No colonies were counted at the lowest level the method reports, here 10 cfu/g. It is a reporting level rather than proof of zero, and it depends on the dilution and volume plated.
Is 10³ cfu/g the same as 1,000 cfu/g?
As a plain number, yes. In pharmacopoeial acceptance criteria such as USP <1111>, 10³ cfu is interpreted as a maximum acceptable count of 2,000, so the two can differ. Check which convention your specification uses.
What are TAMC and TYMC?
Total aerobic microbial count and total combined yeasts and moulds count, the two enumeration results reported under USP <61> and the matching European Pharmacopoeia chapter. Each has its own limit.
What does TNTC mean?
Too numerous to count: the plate carried more colonies than can be counted reliably. It is not a result. Ask for a count from a higher dilution.
Why do food and pharma ingredients have different microbial lists?
The organisms of concern, units and test portions follow the standard and the intended use. Food specifications often name Salmonella and indicator organisms; pharmaceutical ones use pharmacopoeial categories.
Does a passing result mean the whole lot is free of the organism?
No. It describes the portion or portions tested. Contamination can be uneven, so a result is evidence about the sample, and a sampling plan shows how many units it covers.