Ask a laboratory how long a test report is valid and the honest answer is that the report has no expiry date. Nothing in the EMC Directive, the Low Voltage Directive, the Radio Equipment Directive or the FCC rules says a report is good for one year, three years or five. What decides whether an old report still works is not its age. It is three other things - and when someone says "your report has expired", they almost always mean one of them.
The question arrives at a specific moment, usually. A distributor's quality team returns a report you sent three years ago. A marketplace compliance check asks for the current standard version. A supplier form wants reports issued within the last 12 months. Each sounds like an expiry rule. They are not the same rule, and separating them makes the answer much easier to act on.
Three clocks - only one of them is about time
| What you are told | What actually changed |
|---|---|
| "Your report has expired" | The standard edition it cites no longer carries presumption of conformity in the market you are shipping into |
| "This report does not cover the product" | The product changed since testing: adapter, antenna, firmware, PCB revision, accessories |
| "We need a recent report" | The acceptor's own rule - a buyer, marketplace or tender requirement, not the regulation |
Sort a rejection into one of those columns and the work needed becomes obvious: a gap assessment against a new standard edition, a change assessment on the product, or a conversation with the customer about what their rule is actually for.
Clock one: the standard edition
This is the real one, at least in Europe, and it is the reason most reports get refused.
A European standard does not become the reference because CENELEC or ETSI published it. It becomes the reference, in the sense of giving presumption of conformity, because the European Commission has cited it in the Official Journal of the European Union. Testing to a cited harmonised standard is what lets a manufacturer sign a Declaration of Conformity without a notified body going through the file.
When a standard is replaced, the citation tables list a date of cessation of presumption of conformity for the older edition. Between the two editions there is a coexistence period, usually somewhere between 12 and 36 months, during which both versions still carry presumption. After the cessation date, the older edition no longer does. A test report citing that edition still describes real measurements on a real sample - it has not become false - but it no longer supports the presumption you need for products placed on the market after that date.
Two details matter here, and they cut in opposite directions.
The first: publishing a new edition is not by itself the end of the old one. A harmonised standard keeps its presumption of conformity until the Commission sets an end date for it. That has been tested before the Court of Justice of the European Union, which held that a harmonised standard continues to confer presumption of conformity until its date of withdrawal, even after a newer version has been cited. So a customer telling you a report is dead "because there is a new version" may be right - and may be two years early. Find the actual date.
The second: those transition dates are not theoretical. The clearest example in electronics is safety. EN 60950-1 for IT equipment and EN 60065 for audio/video equipment were replaced by EN 62368-1, with the changeover completed at the end of 2020. Reports still citing the older safety standard remain records of real testing, but for a product being placed on the market now they no longer give the presumption the market expects. Radio standards run the same mechanism on their own timetable - the ETSI EN 300 328 and EN 301 489 series are cited and replaced in waves - so if your radio report cites one of those versions, that date is the one to check, not the report's age.
A withdrawal is not a ban, though. Units already placed on the market stay compliant. It is the next production run that has to be assessed against the current edition - which is what Article 7(4) of the EMC Directive asks for when it requires procedures so that series production keeps taking changes in the harmonised standards into account.
Clock two: the product the report describes
A test report is not a statement about a model name. It is a statement about a configuration: hardware version, adapter, antenna, cable lengths, firmware, accessories, PCB revision. The claim inside it reaches exactly as far as that description and no further.
Which is why age is such a poor proxy for relevance. A three-year-old report for a product that has not changed a screw is a stronger document than a three-month-old report for a product whose firmware has been rewritten twice since. The technical file that sits behind a CE declaration is only worth anything if it still matches what is in the box.
A case that comes up constantly: a customer moves to a cheaper power adapter to hold the unit price down, keeps shipping, and keeps quoting the existing report. Nothing about the date is the problem. The report no longer describes the configuration being sold, and the answer is a change assessment on the adapter - which may end in a limited re-test of the affected clauses - rather than a fresh full campaign.
Clock three: whoever is asking
The document rules that actually stop shipments usually are not in the regulation at all. They are in the requirements of whoever is buying from you.
- A distributor's or OEM customer's supplier form may state that test reports must be less than 12 months old. That is a purchasing rule. It is common, it is not law, and it is worth a conversation if the product and the standard have not changed.
- A marketplace compliance check typically looks at which standard version the report cites and which model it covers. Some requests also ask for a recent report, and the check may be run against the current harmonised list rather than the list that applied when you tested.
- An importer or customs officer normally just needs documentation that matches the goods - not a standard-version audit.
- A notified body is a different case again. Where it issued an EU-type examination certificate, read the certificate: some notified bodies issue them with a stated validity period - five years is common - while others issue them with no fixed expiry and tie validity to no change in the product, the manufacturer or the standards used. Both practices exist, and the certificate you hold tells you which one applies.
A certificate issued under a certification scheme can also carry its own validity rules. That is the scheme's rule, not the directive's. A test report is a record of what was measured, and it has no such clock built into it.
Re-issued is not the same as re-tested
When a report is refused, the fastest-looking fix is to ask for it to be re-issued with today's date. Sometimes that is exactly right, because what the acceptor cared about was the date on the document. But a re-issue changes the issue date - not the test date, and not the standard edition cited. If the standard or the product is the real problem, a re-issued report solves nothing, and it puts a document into circulation whose date suggests work that was not done.
| Date on the document | What it actually tells you |
|---|---|
| Date of testing | When the measurements were made, and therefore which rules and standard editions were in force at the time |
| Report issue date | When the document was produced - including a re-issue, which is not the same as a retest |
| Standard reference with year and amendment | The exact edition the product was tested to. This is the field that decides presumption of conformity |
| Addendum or revision number | What has been added since the original: a new clause, a new model, a substituted adapter |
How long do you have to keep the paperwork?
This is a separate obligation from validity, and it is where "how long" has a concrete number.
In the EU, the manufacturer must keep the technical documentation and the EU Declaration of Conformity for 10 years after the apparatus has been placed on the market. The obligation sits in Article 7(3) of the EMC Directive and is mirrored in the Low Voltage and Radio Equipment Directives. The report is part of that file.
In the United States the retention rules run differently. Under 47 CFR 2.938(f), records for equipment authorised under the certification procedure - design drawings, production test procedures, test results - must be kept for one year after marketing of the equipment has been permanently discontinued, and two years for other authorisation procedures. The FCC equipment authorisation itself does not expire; changes to the product are handled through the permissive change rules rather than by re-validating an old grant, which is covered in our guide to whether a product needs an FCC ID or falls under SDoC.
Before you send the report to anyone, check these five things
- Which standard edition does the report cite, including amendments? Compare it with the current OJEU citation list for Europe, or with the applicable FCC procedure.
- Has that edition passed its date of cessation of presumption of conformity?
- Does the tested configuration still match what you build today - adapter, antenna, firmware, PCB revision, rating label, manual, accessories?
- What does the acceptor actually require: a particular standard edition, a maximum age, or simply a report? Ask for it in writing.
- If something changed, is a documented change assessment enough, or do the affected clauses need re-testing?
Those five questions settle the "expiry" question in most real cases, and they take a morning.
So how long is a test report valid?
As long as three things hold at the same time: the standard edition it cites still gives presumption of conformity in the market you are selling into, the product still matches the configuration that was tested, and the party asking for it accepts a report of that scope and age. Age by itself decides nothing. When a report is refused, the reason is nearly always one of those three, and finding out which usually takes one exchange.
Where we come in
Reviewing reports we did not issue is routine work here. This is what a validity and gap review looks like in practice: we take the report you already have, list the standard references it cites, and compare each one against the current citations for the markets you sell into. Then we check the report's product description against what you build today - adapter, antenna, firmware, PCB revision, label and manual - because a report that no longer describes the product will be refused for that reason, whatever its date says.
You get a short written answer: which parts of the file still stand, which clauses are affected, what can be covered by a documented change assessment, and what has to be re-tested. Most reviews come back within a couple of working days - and if the report you hold is fine as it stands, we will tell you that rather than quote you work you do not need. Where a standard version sits outside our scope we say so instead of issuing something that gets questioned later. Our testing and certification scope lists what we run in-house and what we handle with partner laboratories, and our project approach explains how those reviews are scoped and priced.
If a customer, a marketplace or a distributor has questioned an existing report, send us the report, the market where it is being questioned, and a short note of what has changed on the product since it was tested.
We will come back with which parts of the report still hold, which clauses need re-testing, and what a documented change assessment can cover instead.