Request samples by SKU, not by catalog page, and make every unit do two jobs: a display check you can hold in your hands and an information check you can file. That single rule is the working core of a bulk baby toy sample request checklist — one unit, two verdicts, written down before the carton is unsealed. A sample that arrives without the document proving its age grading and test boundary is a toy, not a range decision.
This page covers the papers a customs officer or marketplace reviewer will actually ask for, a worked example of a shipment held for one missing file, and the questions to put to a supplier before you commit to a range slot. It does not cover range gaps or replacing weak sellers — those are separate decisions.
Which documents should be in the sample file before the PO?
| Children's Product Certificate (CPC) — US | Why asked: confirms the children's product was certified against applicable rules. Where it should be: in your sample file before the PO, not after. If you do not have it: ask the supplier for the actual file — CPSC identifies the domestic manufacturer or importer as the certifier and expects details on the test-record holder, manufacture, testing and accepted laboratory. A catalog listing the word 'CPC' is not the record. |
|---|---|
| Third-party test report behind the CPC | Why asked: the certificate is a summary; the report is the evidence. Where it should be: matched to the exact model and configuration on the invoice. If you do not have it: request the report and check the model identity and dates against the sample in your hand. |
| EU technical file and EU declaration of conformity | Why asked: European Commission guidance places conformity assessment, the technical file, the declaration of conformity and CE marking within the manufacturer's responsibilities, with importer and distributor duties alongside. Where it should be: in the product-specific file, not a generic brand folder. If you do not have it: request the product-specific file; a CE symbol in a catalog is not evidence the review was completed. |
| Age grading and intended-use statement | Why asked: age determines which requirements apply — the small-parts ban turns on whether the toy is for children under three. Where it should be: per SKU, on your sample sheet. If you do not have it: do not infer infant or toddler suitability from the category name, the material, or the product photo. Ask for SKU-specific age and intended-use evidence. |
| Button-cell / coin-battery compartment and warning evidence | Why asked: US button-cell rules treat this as a compartment and warning problem, not a toy-function problem. Where it should be: checked on the physical sample — screw closure, tool access, and the warning on product and packaging. If you do not have it: hold the sample. A compartment that opens with fingernails does not ship. |
| Magnet retention and small-parts evidence | Why asked: published recalls turn on magnets that loosen when a cube breaks open, and on accessories that are small parts in an under-three toy. Where it should be: in the abuse-test record for the exact configuration. If you do not have it: ask what evidence shows magnets stay retained and accessories meet the applicable small-parts requirement. |
| Material and component change log | Why asked: CPSC guidance describes retesting affected requirements after a material change and issuing a new CPC using the resulting passing tests. Where it should be: one line per changed material, with the unchanged components listed separately. If you do not have it: request it. Do not assume an earlier report automatically covers the revised configuration. |
| Packaging and display-unit spec | Why asked: display format drives both the retail plan and the shipping and duty calculation. Where it should be: quoted separately so a switch from a display box to a polybag shows up as a number. If you do not have it: ask for the packaging drawing before you approve the sample. |
Worked example (illustrative, not a real shipment)
A buyer consolidates a first range-review order: roughly 4,000 units across four squishy SKUs, split into two destinations — one US, one EU — with the US portion packed in steamer display boxes and the EU portion in plain polybags for a local repack. The sample set arrives complete. The display checks pass: the steamer display stands up on shelf, the print registration is clean, the units are consistent cavity to cavity.
The shipment is held at the US destination exam. The reason is not the toy. The EU-bound portion changed packaging from a display box to a polybag, and the US file still carries the report written against the display-box configuration. Under the material-change principle, an earlier report is not automatically assumed to cover a revised configuration, so the buyer has to go back for the affected requirements and a new certification record. The units are fine. The paperwork is a version behind.
One column on the sample sheet marking changed versus unchanged components per SKU would have caught it, along with a single question before the sample shipped: which requirements are affected by this change, and who coordinates the document review? The hold costs days, not because anyone failed a test, but because nobody wrote down which version was tested.
Which documents do I request with the sample, not after it?
Request the documents at sample stage, because that is the only moment you still have leverage and time. Once the PO is placed and the container is booked, every missing file becomes a delay you pay for.
The practical sequence is short. First, ask for the certificate or declaration file itself — with certifier identity, test-record holder, manufacture details, testing, accepted laboratory, dates and model identity — and compare those details against the physical unit in your hand. Second, ask which specific safety requirements apply to that SKU and whether the named laboratory's scope actually covers each one; CPSC notes that laboratory acceptance is specific to the relevant rules, and one lab may not cover every required test. Third, for EU-bound units, request the product-specific technical file and declaration of conformity rather than a brand-level certificate.
Two details get skipped constantly. The model identity on the report has to match the sample, not the family it belongs to. And the lab scope has to match the configuration, including the packaging change that looks cosmetic but shifts the tested version.
What do I check on the physical sample before I approve the range slot?
Check the hazards the published recalls actually turn on, and check them before you evaluate whether the toy is cute or whether it will sell.
Button-cell and coin-battery units get opened first. Look for screw closure, whether a tool is needed, and what warning appears on the product and the packaging. An electronic baby toy with a compartment a fingernail can lift is a no — the function is irrelevant until that is fixed. US button-cell rules and the door design are the same conversation, just held at your desk instead of at a port.
Then the small-parts and magnet questions. If the SKU includes accessories and the age grading points anywhere near the under-three boundary, the small-parts requirement is the first thing an officer will test against, and a set marketed for the youngest children with loose accessories is exactly the pattern behind published recalls. Magnetic units need evidence that magnets stay retained after the relevant abuse sequence, not a photo of the assembled cube.
For anything with wheels, the wheel and axle retention, rotation, protrusion and small-part checks matter because a loose wheel becomes a hazard after abuse, not before. And for molded parts generally, cavity identification and balance checks are worth asking about — the same machine cycle can still produce cavity-specific differences, which is precisely what a display check on a single sample will not reveal.
How do I keep unsourced numbers out of the sample sheet?
Log every price, MOQ, lead time and test outcome as either sourced or explicitly hypothetical, and leave the cell blank if it is neither. A blank cell invites a question. A guessed number invites a commitment you cannot keep.
This matters most for MOQ and tooling. A mold cannot be assumed transferable merely because a buyer paid a sample or development fee — transfer conditions should be written. If a supplier quotes an MOQ verbally, it goes in the sheet with the source named, or it does not go in at all. Where a corrective action comes up, sorting is containment, not root-cause correction; a plan that addresses the cause and verifies effectiveness is a different document from a rework instruction.
On the supply side, incoming resin inspection should verify supplier, grade, color, lot, packaging condition and required material documents before molding — so when a sample's color drifts from the approved standard, you can trace whether the lot changed.
What goes in the written sample request?
Put seven things in the email and keep it to one page. The SKU list with the display check and the information check named against each line. The changed-versus-unchanged component list. The applicable requirements, with the laboratory scope matched to them. The certificate or file request, spelled out field by field. The SKU-specific age grading and intended-use statement. The packaging and display-unit drawing, quoted separately. And the named person on each side who coordinates the document review.
One illustration of how format and function travel together in a range review: a squishy item such as the Glitter Water-Bead Baozi Squishy with Steamer Display is requested because the steamer display is the shelf fixture, so the sample has to prove the display stands and the units sit consistently — while the information check runs on the age grading and the material documents behind it. The display is the reason you asked; the file is the reason you can ship.
FAQ
How many samples should I request for a range review?
Request at least one unit per SKU for the display check, plus a second unit per SKU where the retail plan depends on display format — one to keep assembled on shelf, one to open. Anything involving a compartment, magnets or abuse testing needs its own unit, because the check destroys it.
Can I accept a catalog page that lists CPC or CE instead of the actual file?
No. A catalog symbol is not evidence the review was performed. Ask for the actual certificate or declaration with certifier identity, test-record holder, manufacture, testing, accepted laboratory, dates and model identity, and match those details to the unit in your hand.
The supplier says the earlier test report covers the new version. Is that enough?
Not automatically. CPSC guidance describes retesting affected requirements after a material change and issuing a new certification using the resulting passing tests. Separate unchanged components from changed materials, ask which requirements the change affects, and confirm who coordinates the document review.
What should I check first on a button-cell baby toy sample?
The compartment, not the toy. Check screw closure, whether a tool is needed to open it, and what warning appears on the product and packaging. US button-cell rules and the door design are the same conversation — a compartment that opens easily is a stop, regardless of how the toy performs.
Does a high rating or reorder volume on a B2B listing tell me anything about sample quality?
No. Ratings and reorder counts describe a listing's history, not the configuration you are about to receive, and they are not verification of test status. Treat them as sourcing leads and resolve quality through the sample and the documents.
How do I confirm a toy is suitable for children under three?
Get SKU-specific age and intended-use evidence from the supplier. Do not infer it from the category name, the material, or the product image — age grading determines which requirements apply, and small accessories in a toy marketed for the youngest children is the exact pattern behind published recalls.
Sources
Request a Quote
Send the SKU, destination market, and the documents you already have.