Honey fraud: what your CoA doesn't prove

Here's an uncomfortable exercise. Pull the certificate of analysis for the last honey delivery your site accepted. Somewhere on it there's probably a C4 sugar result, passed. Moisture, passed. HMF, passed. Diastase, passed. Everything green, supplier approved, intake released, everyone home by five.

Now hold that certificate up against the EU's coordinated "From the Hives" operation, which sampled honey imports at European borders and found 46% suspected of adulteration. Nearly half. Those consignments had paperwork too.

How the fraud actually works

Honey adulteration is not a man with a bucket of sugar water. It's an industrial process run by people who read the same test methods you do.

The classic check — the C4 sugar test (AOAC official method) — works because cane and corn sugars photosynthesise differently from the plants bees forage, leaving an isotopic fingerprint. So the fraudsters stopped using cane and corn. Modern adulteration uses syrups from rice, wheat and sugar beet — C3 plants, isotopically similar to nectar sources — specifically engineered to pass the C4 test. There are syrup manufacturers whose product literature effectively advertises "undetectable by standard honey testing". The adulterant is designed around the CoA.

That's the core problem in one sentence: your CoA proves the honey passed the tests the fraudster designed the syrup to pass.

What the better tests do — and their limits

NMR profiling (and similar broad-spectrum techniques) compares a sample against databases of thousands of authentic honeys, looking at the whole chemical fingerprint rather than one marker. It catches much of what C4 misses. The trade-offs: it costs real money per sample, turnaround isn't intake-friendly, the databases are proprietary and vary by lab, and a determined fraudster with the same instruments is already working on the next syrup. This is an arms race, and the test menu on your spec sheet is always one war behind.

Which is why the labs' own quiet advice is that testing is the last line of defence, not the first.

What this means for your vulnerability assessment

If honey (or anything containing it) crosses your intake, your BRCGS 5.4 / food fraud vulnerability assessment should reflect the actual threat model, not the 2015 one:

Score honey as high-likelihood, by default. A near-50% border failure rate is not an emerging risk; it's the operating environment. If your assessment has honey scored comfortable-green on likelihood, it's describing a different planet.
Interrogate the supply chain, not just the certificate. Country of origin blends ("blend of EU and non-EU honeys" tells you nothing), broker layers, price versus the authentic market rate. Honey priced meaningfully below the cost of producing real honey is answering your question for you.
Make your testing unpredictable. If you test, occasional NMR on a random consignment beats routine C4 on every one — the deterrent is the uncertainty, not the method.
Put authenticity clauses in the spec. Right of audit, declared country-of-origin at drum level, notification of any change in source. Contractual teeth change supplier behaviour more than test schedules do.
Document the reasoning. When the auditor asks why honey is scored high and what you did about it, this analysis is the answer. The assessment that says "reviewed, no change" against a 46% adulteration backdrop is the finding waiting to happen.

The pattern is the point

Honey is the case study, not the lesson. The lesson is that every routine authenticity test eventually gets designed around, because the people adulterating a commodity read the same methods you do and have more financial incentive to study them.

Vanilla: synthetic vanillin passed off as bean-derived, pods dipped in aroma, "natural" declarations that don't survive isotope analysis. Skimmed milk powder: the highest-volume adulteration target in the sector and the one fewest people have on their radar. Herbs and spices: the highest measured adulteration rates of any category — oregano cut with olive leaf, saffron with safflower, paprika with dyes. Meat speciation. Fish substitution. Olive oil, whenever the harvest fails. Cocoa and coffee, right now, because prices have made fraud worth the risk.

Different molecules, identical playbook. Which is why a vulnerability assessment built around a test schedule is the wrong shape — the schedule is always one generation behind the fraud.

The ingredient integrity series on One Step Back: how each fraud actually works, what your testing does and doesn't prove, and what belongs in your assessment — written from the factory side of the desk. Free.

[→ Join the list]