GB 4806 Food Contact Packaging Testing in Practice: 5 Tests That Brands Fail Most + 3 Fixes When You Don't Pass

Last year, in the second half, a client running a frozen ready-to-eat crayfish line sent me a test report marked in red: "Overall migration exceeded." Three retests, all the same conclusion. Her pack was a standard PE vacuum pouch inside an outer paperboard box. Where was the problem? The inks inside the paperboard box outgassed through the mouth of the PE pouch, and the crayfish adsorbed them. The first thing the client asked me was: "Why didn't this show up on the sample I submitted?" It was the second time she had been blocked on the same test.

The GB 4806 standard is written densely, but on a real package only five tests actually get brands stuck. I'll go through them one by one, each paired with a real case — not lifted from the standard text, but cases I have run into over the past three years working with labs, printers, and brand owners.

Overall migration: the silent move from ink to adhesive

Overall migration is a test governed jointly by GB 4806.7 and GB 4806.8. It measures the total amount of non-volatile substances that migrate from the package into the food. What clients most often overlook is that "the inks and laminating adhesives on the outer layers of the package also participate in migration." The crayfish client's PE vacuum pouch itself was compliant, but she had added a sealing sticker across the pouch mouth, and the sticker was UV-printed. The oligomers in the UV ink migrated inward at frozen temperature. That is what GB 4806.7 calls "overall migration" — any contact surface counts, even if there is one layer in between, as long as the path is reachable.

Fix: ask the printer for an "ink contact grade" certificate. Solvent-free UV inks, water-based inks, and EB-curable inks vary enormously in migration; if budget allows, EB-cure is the most stable option.

Primary aromatic amine migration: the forgotten azo dyes

Azo dyes are very common in printing inks because they are cheap and produce highly saturated color. But under acidic conditions they reduce to primary aromatic amines (PAAs), which are carcinogenic. GB 4806.7 lists PAAs as a separate item because of "cumulative toxicity" — the problem is not a single large migration event but small amounts migrating each time and accumulating over the long term.

A client importing baby complementary foods had his box supplier using ordinary four-color ink. He sent samples for testing, and out of 23 PAAs, 3 were detected. A third-party lab retested, still 3. The printer had to rework the job — reprinting with azo-free ink — and three weeks of lead time were already lost.

Fix: if the printer cannot give you a written commitment that "the ink is free of primary aromatic amines," switch printers. Testing costs far more than switching — a full-spectrum PAA test starts at 5,000.

Formaldehyde migration: the hidden pit of lamination and mounting

Formaldehyde migration mainly comes from laminating adhesives and mounting adhesives. GB 4806.7 requires formaldehyde migration ≤ 15 mg/kg. The problem is that many printers' laminating adhesives are based on urea-formaldehyde resin systems, and formaldehyde release exceeds the limit by itself. The "slight smell" clients notice on finished pieces is formaldehyde.

A client running a pastry line had cake boxes made of greyboard mounted with paper, white paper on the inside. The mounting used a water-based laminating adhesive — not a film-lamination process, but formaldehyde was still in the adhesive. His test result was formaldehyde at 18 mg/kg, just over the line. After switching to an EVA-system adhesive, the retest came back at 6 mg/kg and passed.

Fix: any adhesive the printer provides must be labeled "food grade" or "FDA 21 CFR §175.105 compliant." If the printer cannot tell you, ask them to send an adhesive sample for a standalone formaldehyde migration test — far cheaper than retesting the finished product.

Evaporation residue: the real test for liquid-contact packaging

Evaporation residue is the test under GB 4806.7 using a 4% acetic acid simulant, aimed at packaging "in contact with liquid food." The hardest-hit items are straws, plastic cups, and the inner gaskets of sealed lids. A client running yogurt switched to a new supplier's cup lids, and the inner sealing gasket on the lid tested at 32 mg/L evaporation residue against a standard of ≤ 30 mg/L. By the time he called me, he was already at the supplier's office — not to argue, but to ask "can I use the old-design lids we have in stock?" The answer is no. Even if the old-design lids passed before, they are a different batch of compound, so the migration will differ.

Fix: do not cut costs on contact components like straws and sealing gaskets. Food-grade silicone and food-grade PE only cost 20–30% more than the ordinary version, but a single test submission costs from 3,000 — the extra spend pays for itself after three retests.

Heavy metals: four elements, four sources

Lead, cadmium, arsenic, and mercury are the four standard items in the GB 4806 series. Lead mainly comes from ink pigments, especially reds and yellows; cadmium comes from stabilizers; arsenic comes from paper bleaching; mercury almost never shows up in modern processes. The last category, "adhesive backing in indirect contact with plastic," is not a main item for heavy metals but is a common extension trap beyond the 5 categories — a client running honey moved the label from a glass bottle to a plastic bottle, and the plasticizer in the adhesive backing migrated inward through the plastic bottle wall and was detected. Fix: only use self-adhesive labels on glass bottles and tinplate; for plastic bottles, if you must label, use a heat-shrink sleeve.

Lead deserves a closer look. A client running tea had tea cans in tinplate, printed on the outside of the steel. The test report showed lead migration at 0.8 mg/kg against a standard of ≤ 0.01 mg/kg. The problem was chrome yellow pigment in the tin-printing ink. That pigment is cheap and color-stable, but high in lead. After switching to titanium dioxide plus an organic yellow pigment, the retest was 0.003 mg/kg. However, organic yellow pigment has poor lightfastness and will fade within six months — that is a separate trade-off.

Fix: for metal packaging like printed tin and printed aluminum, require "lead-free pigments" in the ink. If you must use a chrome-yellow system, either add an external barrier film or give up chrome yellow.

3 things brand owners can do ahead of time

First, at the packaging design stage, get three compliance declarations from the printer — one each for ink, adhesive, and contact components. The three together cost no more than 500. Second, submit small-batch proof samples for testing first — do not wait until the full production run is printed; the ink batch used for proofing and production may not come from the same drum. Third, retain 3–5 finished samples for at least 2 years, so if you are spot-checked you can trace the batch. I have seen too many brand owners skip these three steps and only start patching things up after three retests.

Compliance for food-contact packaging is not something you can rest easy about after a single test pass. Test reports have a validity window; ink suppliers change; adhesive formulations change. The last piece of advice I can offer: turn your compliance engineer into the printer's friend — that is far more useful than being friends with the lab.

Further reading

What is FDA food-contact material certification? A practical guide to 21 CFR for exporting to the US　How to choose food packaging boxes? A GB 4806 compliance checklist + 5 violation cases　EU REACH certification for packaging materials: how to look up the SVHC list + 5-step process　Compliant use of recycling marks (PCR / recyclable / compostable): 3 mark types + a 4-country regulatory comparison