CE Testing And Certification Laboratory Services In The Eu

Scope and evidence must be confirmed before any paid tests or filings.

A useful CE testing laboratory brief starts before a quotation request. “CE testing” is not one universal test package, and a laboratory report is not a universal “CE certificate”. The required evidence depends on the product, the EU legislation that applies, the conformity-assessment route, the technical specifications used and the product configuration actually placed on the market.

This page is deliberately limited to the testing and evidence decision: what to define before booking a laboratory, how to choose a laboratory by role and scope, how to control samples, when an existing report may remain usable, and what should trigger a retest. For the wider implementation process, see our CE marking services page.

Questions to resolve before booking a CE testing laboratory

Do not begin with “Which lab is cheapest?” Begin with a controlled test brief. The laboratory can only quote the correct scope when the legal and technical inputs are stable enough to identify what must be tested and what evidence the result must support.

Decision	What must be fixed	Why it changes the laboratory scope
Applicable EU legislation	Product function, intended use, electrical/radio characteristics, users and foreseeable environment	Different EU acts can impose different essential requirements and conformity-assessment routes.
Technical basis	Relevant harmonised standards or another documented technical solution, including the edition/date used	A test report is meaningful only against an identified method and acceptance criteria.
Third-party role	Whether the applicable legislation/module requires an independent conformity-assessment body or a notified body	Laboratory competence alone does not give a body a notified-body role for every product or procedure.
Product configuration	Variants, options, firmware, power supply, accessories, critical components and worst-case operating modes	These choices determine representative or worst-case samples and can change the tests.
Evidence destination	Technical documentation, notified-body file, customer acceptance, market-surveillance response or internal engineering decision	The receiving party may require different report format, traceability or accreditation scope.

Using a harmonised standard is generally voluntary, but a cited harmonised standard can provide a presumption of conformity only for the requirements it covers and under the conditions of the relevant EU framework. If another technical solution is used, the manufacturer still has to demonstrate how the applicable legal requirements are met.

Testing matrix: requirement → method → sample → evidence → acceptance

The most useful pre-booking document is a testing matrix. It prevents a common failure mode: buying a test report before anyone has defined which legal requirement the result is supposed to prove.

Matrix field	Question to answer	Booking output
Requirement / hazard	Which essential requirement, risk or performance characteristic needs evidence?	Traceable requirement reference and rationale.
Test basis	Which standard clause, test method or justified engineering method will be used?	Exact document/edition and test clauses.
Sample logic	Which model or variant represents the family, and which configuration is worst case for this test?	Sample list with configuration IDs and reasons.
Expected evidence	Full report, raw data, photographs, component records, calibration references or certificate?	Deliverables written into the quotation or purchase order.
Acceptance route	Who must rely on the evidence: manufacturer, notified body, customer or authority?	Required laboratory status and report conditions confirmed in advance.

For an EMC-specific matrix, use the dedicated EU CE EMC testing requirements page. For low-voltage electrical safety testing, see LVD testing services in the EU.

Select the laboratory by legal role, competence and exact scope

“Accredited laboratory”, “certification body” and “notified body” are not interchangeable labels. Accreditation is a formal attestation of competence for specified conformity-assessment activities. A notified body is designated by an EU country to perform particular tasks under specific EU legislation where third-party involvement is required. Its notified scope should be checked in NANDO, not inferred from a company’s brand or from an unrelated accreditation.

Before appointment, ask for the exact accreditation schedule or scope covering the proposed method, the site that will perform the work, any subcontracted tests, and the report that will be issued. Where a notified body is required, verify the body number, legislation, products/procedures and tasks in NANDO. A laboratory can be technically capable of a test without being the legally appropriate body for a specific conformity-assessment module.

Confirm that the quoted standard, clause and edition are inside the relevant competence/accreditation scope where accredited evidence is required or commercially necessary.
Identify which tests will be subcontracted and whether the receiving body will accept those results.
Ask whether deviations, engineering judgments or partial tests will be clearly identified in the report.
If witnessing or manufacturer-site testing is proposed, confirm who remains responsible for equipment, calibration, method control and report issuance.

Control the sample: representative does not mean convenient

A passing result is only as useful as the link between the tested sample and the production product. The New Legislative Framework’s model modules use representative specimens for type examination and allow the examining body to request further specimens where needed; sector-specific legislation can add or modify requirements. Operationally, the same discipline is useful even where the manufacturer tests without a notified body.

Create a sample identity record before shipment. It should state model and revision, serial or prototype number, PCB/hardware revision, firmware version, critical components, power supply, accessories, cable set, enclosure/material options and the operating mode selected for each test. Photograph labels and internal configuration where appropriate.

For product families, do not assume the highest-rated model is always the worst case. Worst case is test-specific: maximum current may matter for heating; switching frequency and cable arrangement may matter for EMC; enclosure material or spacing may matter for electrical safety. Record the engineering reason for every representative-sample decision.

Can an existing test report be reused?

Reuse is an evidence decision, not a file-format decision. An old report should not be accepted merely because the model name is similar or because the PDF bears an accreditation mark. Compare the tested product, current product, legal basis and intended evidence route.

Check	Reuse may remain defensible when	Retest / additional assessment is more likely when
Product identity	The current product is demonstrably the same relevant design/configuration.	Critical components, layout, enclosure, ratings, radio/electrical architecture or materials changed.
Standard status	The method/edition remains suitable for the applicable requirements and evidence route.	The applicable legal or technical basis changed, or a later edition introduces relevant new/changed tests.
Report completeness	Sample identity, methods, results and deviations are traceable enough to support the technical documentation.	The report is partial, missing configurations/raw evidence needed by the recipient, or its scope is unclear.
Laboratory status	The original laboratory status and scope meet the current receiving party’s needs.	A notified-body procedure, customer or authority requires evidence from a different recognised scope or body.
Product family logic	A documented worst-case/representative rationale still covers the variants.	A new variant creates an untested worst case.

Keep the reuse conclusion in the technical file: what was compared, who reviewed it, what differences were found and why those differences do or do not affect the tested requirements. Where uncertainty is material, obtain the laboratory or relevant conformity-assessment body’s written view before relying on the old report.

Failure and retest: preserve the evidence trail

A failed pre-compliance or formal test is not solved by repeatedly changing the sample until it passes. Use a controlled sequence: record the failure, isolate the cause, implement a documented change, decide what previous results the change can affect, and agree the retest scope before testing again.

Freeze the failing sample configuration and retain the failure data.
Record the root-cause hypothesis and the corrective modification.
Perform an impact review across all relevant requirements, not only the failed clause.
Ask the laboratory which tests must be repeated and whether regression checks are needed.
Link the passing sample and final report to the production bill of materials, drawings and firmware revision.

Partial retesting can be reasonable when technically justified, but the justification should be explicit. A change that improves one test can worsen another: a filter can affect leakage or thermal behaviour; an enclosure change can alter EMC, ingress protection or creepage/clearance conditions. The decision should follow the changed risk, not the cost of the test.

After a pass, use evidence-driven change control

Testing is not finished when the report is issued. The evidence remains useful only while the marketed product remains within the configuration and assumptions that the evidence covers. Maintain a change log that routes relevant design, supplier and software changes through a compliance impact review.

Change trigger	Minimum review question	Possible evidence action
Critical component or supplier substitution	Does the new part change safety ratings, construction, EMC behaviour or certified component conditions?	Document equivalence, targeted test or wider retest.
PCB/layout, power or radio change	Could emissions, immunity, temperature, spacing or protection change?	Engineering review plus relevant regression testing.
Firmware/software update	Does it change safety functions, duty cycle, radio parameters, modes or user safeguards?	Risk review and targeted functional/technical testing.
New enclosure/material/accessory	Does it create a different thermal, mechanical, fire, electrical or EMC worst case?	Representative-sample review and affected tests.
Standard or law update	Does the previous evidence still address current applicable requirements?	Gap analysis; additional tests where the gap is material.

Laboratory booking pack: what to send with the RFQ

A controlled booking pack reduces quotation ambiguity and later change orders. At minimum, provide the product description and intended use, applicable EU acts already identified, proposed standards/editions, model and variant list, critical components, schematics or technical data needed for scoping, operating modes, accessories, sample quantity assumptions, target evidence route, and the exact deliverables required.

Ask the quotation to separate test scope, engineering/review time, repeat tests after failure, sample preparation, travel/witnessing, subcontracted work and optional certification activities. This page does not set market prices; for budgeting logic and cost drivers, use the CE marking cost page.

Before issuing the purchase order, require written confirmation of exclusions and assumptions. Typical ambiguities to eliminate are: “testing to the standard” without clause list, “certificate included” without identifying what certificate, and “accredited testing” without the specific scope/site.

What a laboratory report does — and does not — decide

For many CE-marked products, the manufacturer remains responsible for conformity assessment, technical documentation, the EU declaration of conformity and affixing the CE marking. A laboratory report can be important evidence, but it does not automatically determine every applicable EU act, complete the risk assessment, authorise a declaration, or replace a notified-body decision where the law requires one.

Likewise, “CE certificate” can describe very different commercial or legal documents. Before paying for certification, identify the legal instrument, conformity-assessment module, issuing body, notified scope (if relevant), product/configuration covered, validity or surveillance conditions if any, and how the document fits into the manufacturer’s technical documentation.

Official EU sources to verify the testing route

FAQ: CE testing and certification laboratories

Must CE tests always be performed by an accredited laboratory?

No universal rule says that every test supporting every CE-marked product must always be performed by an accredited laboratory. The applicable EU legislation and conformity-assessment procedure determine whether third-party involvement is required. Even where accreditation is not legally mandatory, an accredited scope may be required by a notified body, customer, authority or internal evidence policy. Confirm acceptance before testing.

Does a passing laboratory report give me a CE certificate?

Not automatically. CE marking is generally based on the manufacturer’s conformity responsibilities, and the role of third parties depends on the applicable legislation. A test report is evidence against identified methods. A certificate issued by a body has to be assessed by its legal basis, issuing scope and the product/procedure it covers.

Can I use a supplier’s or previous model’s test report?

Sometimes, but only after a documented applicability review. Compare product identity, critical components, configuration, standards/editions, laboratory scope and the evidence route. Component or supplier reports may support a file without proving compliance of the complete final product.

Should I do pre-compliance testing before formal testing?

It can be efficient where the design risk is significant or a formal retest would be costly. Pre-compliance results should be clearly separated from final evidence: different equipment, setups or sample maturity can limit how far the results can be relied on.

Which product changes should trigger retesting?

There is no single change list for every CE regime. Trigger a compliance impact review whenever a change can affect an applicable requirement or the assumptions behind existing evidence—for example critical components, PCB/layout, power or radio design, firmware affecting safety functions or operating modes, enclosure/materials, accessories, or a relevant legal/standard update. Retest the affected scope when the review cannot justify continued coverage.