Twenty-six machinery standards stop counting on the same day.

Presumption of conformity is what a designated standard buys you. Build to it, cite it on the declaration, and an enforcement officer starts from the position that the essential requirements are met. It has an expiry date, published in a spreadsheet, and nobody writes to tell you when yours arrives.

So we read thirteen of the consolidated lists at once (machinery, low voltage, construction products, medical devices, PPE, radio equipment, EMC, general product safety, explosives, pyrotechnics, cosmetics, toys, conformity assessment) and asked one question of the whole set: when do these dates actually fall?

Thirteen is not all of them. GOV.UK publishes designated standards for 39 product areas, and the other 26 (fourteen energy-related product types, RoHS, gas appliances, lifts, pressure equipment, measuring instruments and more) are not in any count below.

Why is “only 65 ahead” the wrong thing to be relieved about?

Because 634 have already passed, and every one of those was a live citation on somebody’s declaration the day before it was not.

The number to watch is not how many deadlines exist. It is whether one of them is yours. A firm citing six standards has a roughly one-in-five chance that any given citation carries a date at all, and no way of knowing which without going to look.

What is happening on 20 January 2027?

Twenty-six machinery standards stop conferring presumption of conformity on that single day. Two low-voltage standards go on the 18th and two more on the 23rd. That one month carries 30 of the 65 remaining deadlines: very nearly half.

Clusters like this are an artefact of how the lists are maintained: a batch of standards is superseded together and given a common date. It is entirely reasonable and it is also exactly the shape that overwhelms a testing schedule, because lab capacity does not expand to match a spreadsheet.

Where is the nearest one?

8 October 2026: EN 12941:1998 and EN 12942:1998, both powered respiratory protective devices, on the personal protective equipment list. A declaration signed against either after that day does not carry the presumption it claims.

After that, 15 November 2026 for six radio equipment standards and 16 November for four more on the personal protective equipment list, and then the January cluster.

We read the thirteen lists again on 5 October 2026. None had published a newer version since the 8 September reading (the newest is still machinery, version 16, of 8 August 2026), so every count in this note stands as read. Once 8 October has passed, the 65 still ahead become 63, and the January cluster is 30 of those 63.

The thing almost nobody checks

141 of the 3,517 standards (about four in a hundred) carry a restriction: a note saying which clauses the citation actually covers. The reference on your declaration looks identical whether or not one applies. It is the most-skipped text in the subject and the likeliest single place an audit finds a hole, because nothing about the citation signals that it covers less than the whole standard.

What we could not measure, and why it matters

How much notice you get. We can see that a date exists; we cannot see from one reading when it appeared. A standard whose end date was published three years ago and one whose end date was added last Tuesday look exactly the same in today’s spreadsheet.

That is not a gap we can close by reading harder. It closes by keeping yesterday’s copy and comparing, which is the entire argument for monitoring a public dataset rather than checking it. A spot check finds the reference you thought to look up. The risk is the one you did not.

What to do with this

1. Check the references you actually cite, not the standards you build to. Presumption attaches to the reference as written on the declaration, edition and amendments included: the checker runs on this same corpus and takes ten seconds.
2. If you make machinery, look at January 2027 first. Twenty-six on one day is a scheduling problem before it is a compliance one.
3. Work back from the date, not forward from today. Buying the successor standard, re-testing and re-issuing is the long pole, and it does not compress.
4. Read the restriction where there is one. Four per cent of citations cover less than the standard they name.
5. Do not assume the EU agrees. Great Britain publishes designated standards; the EU publishes harmonised ones in the Official Journal. They were the same lists once. They are not now.

None of this is a conformity assessment, and none of it is advice about your product. It is what thirteen government spreadsheets say, read on one afternoon, with the arithmetic shown.