Preservative challenge testing under the Chinese standard: what day 7, 14 and 28 tell you
The preservative challenge test is the one test that fails quietly: the product looks fine, smells fine, and is still not safe. Here is how it runs under the Chinese standard and what usually goes wrong.
The short answer: the product is inoculated with a defined set of organisms, sampled at day 7, 14 and 28, and the log reduction in each organism is compared against the standard’s acceptance criteria. It tests the formula as a system — preservative, pH, water activity and packaging together — not the preservative alone. Stability testing is the other half of that picture.
Why it exists
A cosmetic is not sterile and does not need to be. It needs to resist becoming a culture medium once a customer opens it, dips fingers in it, and leaves it in a warm bathroom.
The challenge test simulates that abuse deliberately. If the formula can neutralise a heavy inoculum and keep it neutralised, it will survive normal use. If it cannot, no amount of good manufacturing practice will save it after opening.
How the test runs
- Inoculation — the sample is challenged with a panel of organisms covering Gram-positive bacteria, Gram-negative bacteria, yeast and mould. The standard specifies which.
- Sampling — aliquots are taken at day 7, day 14 and day 28, and the surviving count is measured for each organism.
- Judgement — the log reduction at each interval is compared against the acceptance criteria in the standard, and the product is graded accordingly.
The reason the test runs 28 days rather than 7 is that some preservative systems look strong early and fade. A system that holds at day 7 and collapses at day 28 is a system that fails in a customer’s third week of use.
The three habits that make formulas fail
1. Formulating for the label, not the system. Adding a "preservative-free" claim by removing the preservative and relying on a multifunctional raw material. Sometimes it works. Often it passes day 7 and fails day 28.
2. Ignoring pH. Most preservative systems have a narrow effective pH band. A formula buffered to a pH where the preservative is largely undissociated and inactive will fail regardless of how much was added.
3. High water activity with high nutrient load. Botanical extracts, proteins and sugars are excellent marketing and excellent microbial food. A formula loaded with them needs a stronger system, not the same one at the same dose.
China’s standard versus ISO 11930
The Chinese cosmetics safety standard and ISO 11930 share the same logic — inoculate, sample over time, grade against criteria — but they differ in organism panel, sampling intervals and the numeric thresholds for each grade.
What this means practically: a product tested to ISO 11930 for the European market is not automatically compliant with the Chinese requirement, and vice versa. If you are selling into both, ask which standard the report was issued against. A report that does not name the standard is not usable for either.
What the report should contain
- The standard it was tested against, named explicitly
- The organism panel used, with strain references
- The counts at each sampling interval, per organism
- The grade assigned, and the criteria it was graded against
- The formula code and batch tested — not just the product name
If any of these are missing, the report cannot support a safety assessment in any regulated market.
FAQ
Does every product need a challenge test?
Any product containing water does. Anhydrous products — balms, oils, pressed powders — generally do not, because they cannot support microbial growth.
How long does it take?
The test itself is 28 days. Allow time for the report to be issued afterwards.
Can I reuse a challenge test from another product?
Only if the formula is identical. Changing a preservative, a pH adjuster or a botanical extract changes the answer.
What if my product fails?
The system is adjusted — preservative type, dose, pH, or packaging — and the test restarts. It is a formulation problem with a formulation solution.
Related reading
- Stability testing at -25°C and 60°C: what a three-month run is looking for — Why we test at both extremes rather than a comfortable middle, what fails first, and what passing stability still does not prove.
- Serum manufacturing: why viscosity is different for every formula — Viscosity is not a setting on the filling line.
- How to choose a skincare manufacturer in China: a 2026 buyer’s checklist — Licence, unit scope, stability data, documentation and audit access — the five things that separate a real factory from a trading desk.
Sourcing this? Quality & Compliance · Send a brief · FAQ