TL;DR: A Certificate of Analysis that doesn’t list specific migration limits, test method references, and lot traceability is not a compliance document — it’s a decoration.
TL;DR: In our incoming inspection protocol, we reject ink lots where heavy metal content exceeds 90 ppm total extractable, even when the ASTM F963-17 threshold sits at 600 ppm — our internal buffer accounts for print laydown variability across different substrate absorbencies.
What Failing Compliance Documentation Looks Like in Practice #
Three symptoms show up repeatedly when a brand brings us packaging from a previous supplier that has failed toy safety compliance review:
Generic COA language with no lot linkage. The certificate says “complies with EN 71-3:2019” but carries no batch number, no test date, and no reference to which migration category applies. Category III colorants have a soluble chromium limit of 1.3 mg/kg under EN 71-3 — a COA that doesn’t specify which category was tested tells you nothing usable.
Missing substrate-specific data. A COA issued for a paper liner is presented as covering the entire package, including a coated board outer shell with UV-cured flood varnish. The two substrates have different migration profiles. A single document covering both is not valid under EN 71-3:2019 Article 7.2 testing scope.
Test reports from labs that don’t hold ISO/IEC 17025 accreditation. Under ASTM F963-17 Section 4.1 and the parallel EU General Product Safety Directive requirements, test data used for regulatory substantiation should come from accredited facilities. If your supplier’s test report comes from a non-accredited internal lab, you have a documentation gap, not a compliance confirmation.
| Documentation Red Flag | What It Usually Means | Risk Level |
|---|---|---|
| COA with no batch/lot number | Document may be reused across multiple production runs | High |
| Single COA covers mixed substrate types | Substrate-specific migration not individually confirmed | High |
| Heavy metal values listed as “<600 ppm” without method | Test method (ICP-MS vs. AAS) and detection limit unknown | Medium |
| No ISO/IEC 17025 lab reference | Test data may not be defensible in regulatory review | High |
| COA issued by supplier’s own QC department only | Third-party verification absent | Medium |
| “Pass/Fail” results only, no numerical data | Cannot assess margin to limit; no early warning signal | Medium |
The Migration Pathway Problem Most Qualification Audits Miss #
The root cause that gets misdiagnosed most often in toy packaging compliance is the confusion between total heavy metal content and extractable (soluble) heavy metal content. These are different measurements with different regulatory relevance, and a COA showing one does not satisfy the requirement for the other.
EN 71-3:2019 specifies migration limits in mg/kg of dry material, measured under a simulated gastric fluid extraction (0.07 mol/L hydrochloric acid, 1 hour at 37°C per ISO 8124-3 annex methodology). ASTM F963-17 references a similar extraction approach under its Section 4.3.5.2 requirements. What matters is not how much lead is in the ink film — what matters is how much lead can be extracted under those conditions and therefore ingested by a child.
A supplier who tests for total metal content using X-ray fluorescence (XRF) is doing a screening scan, not a compliance test. XRF gives you elemental totals. It cannot distinguish between a tightly bound pigment lattice where metal is essentially non-extractable and a soluble colorant where the same metal leaches readily under acidic conditions. PB29 (Pigment Blue 29, ultramarine) contains aluminum and sulfur but is essentially inert under gastric conditions. Contrast that with chrome yellow pigments, which are highly soluble and now banned from toy-compliant inks under EN 71-3 Category I limits.
The confirmation threshold we use in our incoming inspection: if a supplier presents XRF data only, we request ICP-MS (inductively coupled plasma mass spectrometry) data on any ink or coating applied at >3 g/m² on food-contact-adjacent or toy-category substrates. Below 3 g/m² laydown, XRF screening is acceptable as a first-pass filter only — not as release documentation.
Measurement protocol from our QC-14 incoming ink assessment form: a minimum of 3 ink drawdowns on the actual production substrate are tested per lot. The drawdowns are done at the production-equivalent laydown rate, not at a laboratory test thickness. This matters because migration is partially a function of film thickness, and labs that test at standardized thickness may underreport migration from high-laydown production conditions.
Corrective Actions When a Supplier’s Qualification Documentation Fails #
-
Request ICP-MS re-testing at an ISO/IEC 17025-accredited lab, at your expense if needed. This fixes the documentation gap for the current lot and gives you baseline numbers for future comparison. Turnaround from a qualified lab runs 7–10 working days for standard heavy metal panels, 15–20 days for full EN 71-3 Category II/III or ASTM F963-17 Section 4 panels. This resolves the immediate compliance question but doesn’t address the supplier’s process capability.
-
Issue a Corrective Action Request (CAR) specifying COA field requirements. A compliant COA for toy packaging ink or board should include: lot number, production date, substrate type, test method reference (ICP-MS or AAS), individual element values in mg/kg, applicable regulatory limit for each element, name and accreditation number of testing lab, and authorized signatory. If a supplier cannot produce this within 10 working days, that’s a process maturity signal worth weighing in your sourcing decision.
-
Audit the supplier’s incoming raw material controls. Non-compliant finished packaging usually originates from a pigment or coating that wasn’t verified at goods-in. Ask to see the supplier’s approved vendor list (AVL) for inks and coatings and their incoming lot-release criteria. If they can’t show you written pass/fail thresholds for heavy metal content on incoming pigment lots, their compliance program is reactive, not preventive.
-
Switch to compliant ink chemistry, not just compliant test results. For toy packaging that requires ongoing compliance without per-lot testing overhead, specifying water-based inks formulated against the toy safety positive lists (e.g., EuPIA Good Manufacturing Practice guidelines) shifts the control upstream. Our sustainable packaging line uses water-based inks on all children’s product packaging by default for this reason.
-
Implement periodic third-party surveillance testing, not just qualification testing. Pigment formulations change. A supplier who passes qualification in Q1 may switch raw material sources in Q3 without notification. Annual surveillance testing per ASTM F963-17 Section 1.5 lifecycle provisions is the minimum we’d recommend for any active toy packaging SKU.
What to Specify Upfront to Prevent Qualification Failures #
Put these requirements in your supplier brief before sampling begins, not after a failed audit:
- Specify “EN 71-3:2019 Category [I/II/III] compliant” on the PO line item, not just “toy safe”
- Require ICP-MS test reports, not XRF screening, for all inks and coatings applied to finished packaging
- Name the required accreditation: ISO/IEC 17025, with lab name and accreditation number on all COAs
- Require individual element values, not pass/fail summaries
- Set a lot traceability requirement: COA batch number must match production lot label on each carton
The document to request before approving any new supplier: their full incoming inspection procedure for ink and coating raw materials, including written pass/fail thresholds. If they don’t have a written procedure, they don’t have a repeatable compliance process.
Specification Notes for Brand Partners #
When you brief us on toy-category or children’s product packaging, the single most important piece of information you can give us upfront is the regulatory market: US (ASTM F963-17), EU (EN 71-3:2019), or both. The element-specific limits differ, and the test methodologies have subtle divergences that affect which ink systems we specify.
A gap we see repeatedly in incoming briefs: the brand specifies “non-toxic inks” without indicating whether the product is for children under 36 months. Age band matters because EN 71-3 Category I limits (for materials that can be scraped off or mouthed by infants) are significantly tighter than Category III limits that apply to large, fixed surface-area components.
Our standard sampling timeline for toy-compliant packaging is 18–22 working days, which includes 7–10 days for third-party ink and board compliance testing before we release samples for your review. If you need samples faster, we can run structural samples on non-production stock for fit/form review while compliance testing runs in parallel — but we don’t release production-representative samples without test data in hand.
Does a COA from my current supplier cover our packaging automatically?
Not unless it references the specific substrate, ink lot, and test method used on your actual production run. A COA issued for a different substrate or from a different production period is not transferable.
We’re selling in both the US and EU — do we need separate test reports?
For most elements, yes. The extractable antimony limit under EN 71-3:2019 Category II is 560 mg/kg; ASTM F963-17 sets a 60 ppm total antimony limit on substrate materials. The methodologies and thresholds don’t map cleanly onto each other, so dual-market products typically need parallel test programs.
Can we skip third-party testing if our supplier provides a self-declaration?
A self-declaration satisfies some administrative requirements under the EU General Product Safety Regulation (GPSR), but it doesn’t replace test data for the inks and coatings specifically. If a market surveillance authority pulls your product, a self-declaration without underlying test reports is not a defensible compliance position.
Our packaging has a spot UV coating — does that need separate testing?
Yes, and this catches brands off guard. UV-cured coatings can contain photoinitiators that are subject to migration restrictions under both EN 71-3 and food-contact frameworks. If the UV varnish is applied over a printable surface that a child could mouth, it needs its own COA referencing the specific cure energy (we typically run at 120–160 mJ/cm² for full cure on toy packaging) and the photoinitiator content of the varnish formulation.
What’s a realistic MOQ for toy-compliant folding cartons with full documentation?
Our standard MOQ for toy-category folding cartons with third-party compliance documentation is 5,000 units per SKU. Below that threshold, the per-unit cost of compliance testing (which runs $350–$600 per full EN 71-3 panel at accredited labs) starts to outweigh the production savings from Chinese OEM sourcing, and we’d flag that to any brand running a small initial order.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
Switching to uncoated kraft for our inner trays last year actually simplified the EN 71-3 substrate documentation headache — one material, one migration profile, one COA scope. The UV flood varnish on coated board was the exact scenario that kept tripping us up with Category II colorant limits, separate test runs for each substrate layer adds weeks to qualification cycles.
The ISO/IEC 17025 point is mostly solid, but there’s a carve-out worth flagging — during the transition window after a lab’s accreditation lapses for renewal (we had a 47-day gap with our primary test facility in Q3 last year), some regulatory bodies will accept data from that facility if the scope of accreditation was valid at the time of testing. We didn’t lose a single submission over it, but only because we documented the accreditation status on the test date specifically, not the submission date.
We’ve had COAs come in from two different print runs on the same job — same document, different lot numbers hand-corrected in pen — and both claimed EN 71-3:2019 Category II compliance with no method cited. Getting substrate-specific retests from an ISO/IEC 17025 lab added six weeks to our Q4 launch window, which nobody budgets for.
We flag any COA where chromium is reported as a single value without specifying the Category II vs. Category III split — had a supplier pass their own review using Category II limits on a colorant that was clearly Category III, and the 1.3 mg/kg threshold difference didn’t surface until our lab re-ran it under EN 71-3:2019 Annex B.
Had a heat-seal failure on a 200,000-unit blister card run for a handheld gaming accessory line — the foil spec we’d approved was 38-micron PET/foil laminate, but the supplier had quietly substituted 30-micron on the second production pull without updating the COA. The COA still referenced the original material spec, passed incoming inspection, and we didn’t catch the switch until retail returns started spiking around week 8. That substrate change also invalidated our EN 71-3 migration data entirely, since the original test profile was substrate-specific — one undocumented material swap and six months of compliance documentation was effectively decoration, exactly as described here.
Switching to molded pulp inserts for a small electronics accessory line last spring forced us to re-qualify every ink vendor from scratch — pulp’s absorption variability meant our existing COAs, all issued against coated board, were worthless for migration substantiation under EN 71-3:2019 Article 7.2, and three of our five suppliers couldn’t produce substrate-specific data at all.
We started requiring suppliers to include the actual numeric migration values on every COA, not just a “complies with EN 71-3:2019” checkbox — caught a chromium result of 1.1 mg/kg on a coated outer shell last quarter that would’ve sailed through on a blanket pass/fail statement, and that’s uncomfortably close to the 1.3 mg/kg Category III limit given our print laydown variance.