Dating a watch

Not one of 395 watch makers publishes a serial-to-year table. Not Rolex, not Omega, not Seiko, not Patek. Ten other families have been swept for this site and several had at least one maker who published their figures. Watches have none.

What several of them do instead is sell you the answer. That is the thing worth knowing before you start, and it is unique to this trade: in watchmaking the production date is a service with a price on it, not a reference table.

The date is a product here

Illinois Watch are the clearest case because their price is on the page. You send the serial number, and they go and look: “our experts will go through the Illinois Watch records to provide you all the information available for your watch”. The records exist. They are simply not published, and the labour of reading them is what you are buying.

What we could not read, and will not guess at. Omega's Extract of the Archives page and Rolex's WatchID both render entirely in JavaScript, so a fetch returns navigation and nothing else. We can tell you those services exist, because the makers link to them from their own customer-service pages. We cannot tell you what they cost or what period they cover, and we are not going to estimate.

The ones that confirm the date is in there, and stop

Citizen say their serial number reflects the manufacture date — and do not publish which digits mean what. That is a maker confirming a scheme exists while keeping it. Their support site returned HTTP 403 to us, so we could not read further, and when a site will not tell us its rules we stop asking.

Several others publish a page explaining where the serial number is engraved — on the caseback, on a lug, inside the case on vintage pieces — and nothing about reading it. Frédérique Constant is typical. Knowing where to look is genuinely useful and is not the same as knowing what it says.

So what about the tables you can find elsewhere?

They exist, they are often good, and they are somebody's unpaid work — collectors who have logged thousands of watches over decades and published the result. Some of it is also sold in books.

None of it is reproduced here. This site copies a manufacturer's own published figures, with attribution and a permission letter, or it copies nothing. A collector's compilation is not the maker's data, and a serial table lifted from an enthusiast's site is exactly the thing this site declines to do — the guitar page removed 103 Gibson ranges for the same reason, and the piano page will not retype the Pierce Piano Atlas.

If you want those tables, go to the people who made them. They earned the traffic.

Rolex, and why the serial may not answer at all

Rolex is the single most-searched serial number on this site's whole keyword list. It is also the one where a table would help you least, for a reason that has nothing to do with who publishes what.

Rolex publish no table, so every chart you will find is somebody's reconstruction from watches they have seen. That has one consequence worth stating plainly, and it holds whatever Rolex's numbering practice has been: a reconstruction can only cover what its compiler observed. Each chart therefore ends somewhere, and a serial past that end is not a puzzle to solve — it is the chart telling you it does not cover your watch.

We are not going to tell you how Rolex have numbered watches in different eras, because we have not read that from Rolex. It is widely written about elsewhere; none of it comes from them, and this page does not repeat claims it cannot source.

This page does not carry a Rolex chart, so the question does not arise here. It is worth saying because the charts elsewhere rarely say it themselves.

What was checked

The pool is 395 watch makers, listed in watch-makers.txt, from Wikipedia's watch brand, Swiss watch brand and watch manufacturing categories, with websites from Wikidata's official-website property rather than guessed.

And read the swept figure knowing how much was there to read. Of the 181 counted as swept, 36 offered this method fewer than 25 pages to look at, and 33 of those offered nothing from a sitemap at all. A host that answers, allows us, and then has almost nothing to search still counts as swept. It is the number of doors opened, not the number of rooms searched.

Of the 108 not probed: 51 did not answer, 20 returned 403 on robots.txt itself, 2 disallow us in robots.txt, and 12 went over our 45-second budget for a single host. Among those refusing outright: Bulgari, Casio, Chanel, Fossil, Girard-Perregaux and H. Moser & Cie. Nothing was taken from any of them.

That 89 is the highest refusal rate of any family here — more than a quarter of the pool — and luxury brands are noticeably over-represented in it. Read 151 as swept, not as cleared: this method has demonstrably missed makers who do publish, on Zippo, Ludwig and Rheem.

Building the list

A watch maker list built from Wikipedia categories collects a remarkable amount that is not a watch maker. Seventy-five entries were removed, among them Wearing My Rolex (a 2008 single), Rolex on a Redneck (a 2021 song), the Rolex Masters, the Rolex Learning Center, Rolex Tower, Rolex v EUIPO, and a 1998 murder case.

They go by a positive test: a company is filed under a category containing “companies”, “manufacturers”, “corporations”, “businesses” or “brands”, and anything with none of those is dropped and reported. Blacklisting categories instead is how an earlier list nearly lost Sanyo and Toshiba to the word film. All twenty-five of the major brands survived the filter.