Check a Chinese Supplier's Product Test Report

The report was real. The conclusion was passing. The product in the buyer's order had never been tested.

A buyer preparing an order for 12,000 insulated drink bottles received a polished laboratory report from the supplier. Its number resolved to a real institution, and the result for the tested material met the stated limit. Several pages later, however, the sample description identified only a flat stainless-steel coupon submitted by a related material trader. The report did not identify the finished bottle, lid coating, gasket, weld, model, factory, or current bill of materials.

The document was useful evidence about one piece of material. It was not evidence that the assembled product had passed every test the buyer needed. That distinction is the center of a sound review: first establish whether the report is authentic, then determine exactly what was tested, under which method, and whether that evidence still connects to the product that will be manufactured.

Define the claim before opening the PDF

Write a one-sentence requirement that is independent of the supplier's document. Name the product, destination, model or SKU, relevant materials and components, production configuration, and the decision the report is meant to support. If the test program is regulated or customer-mandated, identify the applicable requirement and edition with a qualified compliance owner before reviewing evidence.

Decision to support: whether bottle model VB-24, in the quoted production configuration and all three colour variants, has suitable evidence for the specified material and finished-product requirements for the destination market.

This sentence prevents a common reversal. Without it, the reviewer starts from a supplier's passing report and gradually changes the requirement to fit whatever the laboratory happened to test. A metal-composition result becomes “food-contact tested”; a component report becomes a finished-product report; a report for one model is stretched across a catalogue.

Also identify the document type. A laboratory test report records tests and results for described samples. It is not automatically a product certificate, regulatory approval, declaration of conformity, factory audit, or shipment inspection. For CE-marked products, the CE documents review begins with the applicable legislation and manufacturer obligations; one laboratory report may support that file but cannot replace the whole legal route. An ISO 9001 certificate check answers a different question about a quality management system.

Turn the complete file into a report identity sheet

Ask for the original complete PDF, including appendices, photographs, blank-looking continuation pages, amendments, and any model schedule. A cover page or marketplace screenshot is not enough. Preserve the supplier's email and file as received; work from a review copy rather than overwriting it with comments.

Record the following fields without interpreting them yet:

• report number, revision or amendment number, and total page count;
• laboratory legal name, testing site, contact details shown, and any registration or accreditation number;
• applicant, customer, manufacturer, factory, brand owner, and their stated roles;
• sample name, model, markings, materials, quantity, condition, photographs, and who supplied or selected it;
• sample receipt, test, issue, and amendment dates;
• standards, methods, editions, deviations, test items, limits, measurements, uncertainties where relevant, and conclusions;
• subcontracted or externally supplied results, customer-provided statements, signatures, seals, and accreditation claims.

Do not force the applicant and manufacturer to match. A trader, brand owner, importer, component producer, or laboratory customer may legitimately be the applicant. The report becomes misleading only when a commercial claim gives that role a meaning the report does not support. Keep the role map factual: who paid for the test, who made the sample, whose model appears, and which entity will make the ordered goods.

Page integrity matters. Confirm that the report number, page numbering, footer, dates, and revision status remain consistent from beginning to end. An appendix from another report, an unexplained replacement page, or results without the referenced sample photographs needs issuer confirmation. It is an unresolved discrepancy, not automatic proof of fraud.

Verify the record and issuer through independent routes

For a report within the Chinese public query system's coverage, start at the CNCA official business-systems directory and follow the entry for inspection and testing report-number queries. The CNCA's platform explanation says a report number can return fields such as the issuing institution, its address, qualification certificate, and report date.

Compare those fields with the PDF rather than treating “found” as the conclusion. A database match does not show that every page in the supplier's file is unchanged, that the sample description is commercially accurate, or that the tested item is today's product. A missing result also needs proportionate handling: confirm the number and punctuation, ask whether the report falls within the platform's coverage, and continue with the issuer before deciding not to rely on it.

Find the laboratory's website and contact details independently. Do not use only a QR code, email address, or telephone number printed in a disputed file. Send the full report, or at least the number, issue date, applicant, sample, and page count, and ask the laboratory to confirm that it issued that revision. Ask separately whether later amendments, withdrawals, or superseding reports exist.

If the report displays a CNAS symbol, the CNAS verification FAQ directs users to check the institution through CNAS information services and to contact the laboratory about report authenticity. Save the search terms, access time, result, and laboratory response. This makes the check repeatable when a buyer, auditor, regulator, or colleague asks how the file was accepted.

Check the laboratory's scope, not just its symbol

ISO describes ISO/IEC 17025:2017 as the international standard for the competence, impartiality, and consistent operation of testing and calibration laboratories. Accreditation bodies use it as criteria when assessing laboratories. ISO does not accredit the laboratory, choose the sample, or approve the buyer's product.

Use the CNAS official accredited-laboratory directory to match the laboratory's legal identity and registration. Then open the recognized capability scope and compare the actual test method, product or material category, parameter, laboratory site, and relevant dates. A laboratory can be accredited and still issue a result outside its accredited scope. The logo alone cannot answer that narrower question.

For reports issued elsewhere, the ILAC guidance on accredited conformity assessment bodies points users to the responsible accreditation body's directory and scope. International recognition can support acceptance of competent results, but it does not decide whether a destination regulator, customer specification, certification scheme, or contract requires a particular laboratory or test route.

Record the outcome precisely: laboratory identity verified; claimed accreditation active; listed method within scope, or name the element that remains unresolved. Avoid vague notes such as “CNAS approved” or “ISO certified lab,” which collapse several separate checks and can be misunderstood later.

Read the sample description before the PASS line

The current Chinese inspection and testing supervision rules draw a useful boundary for client-submitted samples: the report is responsible for the tested items' conformity, while the client is responsible for the submitted sample's representativeness and authenticity. That boundary is easy to lose in a supplier sales file.

Ask who selected the sample and from where. Was it taken independently from routine production, selected by the buyer from a controlled sample set, or prepared and mailed by the supplier? Was it a production unit, prototype, component, raw-material coupon, or a specially built “golden” specimen? A laboratory can test a specimen competently without knowing whether it represents the future shipment.

The CNAS reporting clarification CNAS-EL-13 uses an instructive model-family example: when only one model is actually tested, other customer-claimed models should not simply be presented as tested unless the laboratory verifies the relationship and records supporting information. It also distinguishes customer-provided information, off-site testing, witnessed testing, and externally sourced report content.

Return to the bottle report. The photograph shows a rectangular metal coupon. The sample name is a stainless-steel sheet grade, and the applicant is a material trader. The ordered bottle has an inner wall, outer wall, weld, coating, lid, gasket, and printed decoration. The result may support the stated composition of the tested coupon. It does not establish the grade of every production part, migration from coatings, gasket properties, finished-product construction, or consistency across 12,000 units.

When the sample route is weak, do not rewrite the laboratory's conclusion. Narrow the procurement conclusion instead. For a fresh test, the buyer can improve control by defining the sample, selecting it from identified units, recording serial or lot information, sealing it where practical, and keeping photographs and dispatch evidence. The product sample workflow explains how to preserve a controlled approved sample before volume production.

Follow every method into the result pages

“PASS” is shorthand for a comparison made under stated conditions. Identify the test method and edition, the requirement or limit applied, any deviations, and whether the method suits the material, product, configuration, and destination decision. A familiar standard number with the wrong year or an unrelated part of a standard may not answer the buyer's requirement.

Read each required item, not only the overall conclusion. Mark items that are absent, not applicable, not tested, client-declared, subcontracted, witnessed, estimated, or reported without a conformity judgment. Confirm that the measured result and unit can be compared with the stated limit. When a report combines data from different dates or revisions, determine which sample and configuration produced each result.

The Chinese supervision rules require disclosure of subcontracted testing within their scope and address signatures, institutional seals, technical bases, and corrections. A subcontract is not inherently a problem; an undisclosed or unverified one breaks traceability. Identify the organization that performed the item, then check its competence and the relationship between its result and the main report.

Retesting only failed items also needs context. If an earlier sample failed, the supplier modified it, and a later report tests only the failed items, do not present the later pages as though one unchanged specimen passed the entire program. Preserve the original failure, the modification, the retest scope, and the evidence connecting the unchanged passing items to the final construction.

Reconnect the report to the product that will be made

A valid historical report becomes procurement evidence only through configuration control. Compare the report sample with the quotation, approved drawing, bill of materials, material specifications, supplier part numbers, product photographs, labels, firmware or software version where relevant, factory, production process, and change history. The comparison should be specific enough that another reviewer can see why the old result does or does not apply.

Build a short delta list. For the bottle, it might say: tested stainless-steel coupon grade matches the declared inner-wall specification; outer wall supplied by a different mill; weld not represented; coating, lid, gasket, and printed decoration not tested; finished model absent from report; factory connection unverified. This is more useful than a red or green document score.

Ask the laboratory for a model-family or construction rationale when a supplier claims that one result covers variants. Do not invent equivalence from similar photographs or a shared marketing name. Colour may be irrelevant to one mechanical test but decisive for a coating or chemical test. A power supply change may leave enclosure dimensions unchanged while altering electrical compliance. Relevance is test-specific.

A report cannot prove that the quoted factory controls the tested construction in routine production. Follow the manufacturer verification workflow when site, process, subcontracting, equipment, capacity, or quality records remain material. Later, a pre-shipment inspection can compare an identified production lot with approved specifications, but it should not be described as a substitute for laboratory tests it cannot perform on site.

Give the report a bounded procurement disposition

Disposition	What the evidence supports	Procurement action
Reusable evidence	Issuer and report confirmed; competent scope; suitable methods; complete results; tested sample and current configuration demonstrably match.	Record the supported claims and keep the configuration under change control.
Limited evidence	The report is authentic and useful for one stated material, component, model, or test item, but not the supplier's broader claim.	Accept only the narrow finding and list every uncovered element.
Delta testing needed	A sound base report exists, but a changed component, new variant, later method, or missing item affects the current product.	Have a qualified owner define targeted testing and the model-family rationale.
Unresolved; do not rely	Issuer, pages, scope, sample, method, result, or production connection cannot be reconciled.	Pause the affected approval and obtain issuer confirmation or commission controlled testing.

The bottle report receives limited evidence. Its verified result can stay in the file as evidence about the tested coupon. The buyer still needs a destination-specific assessment of the finished product, a controlled construction record, and testing for the gaps identified by the qualified owner. No accusation is required, and no passing claim is expanded beyond the pages.

Commission the next test so it answers the buying decision

When existing evidence is insufficient, send the laboratory a test brief rather than asking for “a full test.” State the destination and decision, exact models and configurations, applicable methods and editions, required items, sample quantity, sample-selection owner, prototype or production status, reporting language, model-family assumptions, acceptance of subcontracting, and the form of results needed. Ask the laboratory to identify conflicts or missing information before testing begins.

Keep evidence of sample selection and dispatch. Require the report to identify the sample actually received and any customer-provided fields. Agree how modifications, retests, partial tests, deviations, and amendments will be shown. Define change triggers for materials, components, suppliers, factory, tooling, process, firmware, design, and test requirements. High-consequence or regulated products need a competent specialist and the relevant authority's current rules; a general supplier guide cannot select that legal route.

Finally, place a dated report-use memo in the supplier approval file. Record the report and revision, sources checked, laboratory and scope result, tested sample, methods and items, configuration comparison, supported claims, exclusions, open gaps, owner, decision, and refresh triggers. That memo is the usable output. “Report verified” is not.

This guide provides general commercial due-diligence information. It is not laboratory accreditation, product certification, regulatory, engineering, or legal advice.