How Chronometer Certification Works: COSC, METAS, Rolex and Beyond

COSC is the Swiss official chronometer testing institute, and its standard follows ISO 3159. A mechanical movement is tested over roughly two weeks, in several positions and at three temperatures (8, 23 and 38 degrees Celsius). To pass, the average daily rate has to land between -4 and +6 seconds per day.

The detail owners often miss is that COSC tests the movement before it's cased. The certificate says the caliber can keep that rate on a test bench. It doesn't promise the same result once the movement sits in a case, on your wrist, with a crown and gaskets and a rotor spinning against your daily routine.

That's still a meaningful bar. A COSC certificate means someone other than the brand checked the movement, which is why it remains the baseline across most Swiss luxury watchmaking.

Yes. In February 2026 COSC announced an Excellence Chronometer designation. A movement is first certified as a standard chronometer, goes back to the brand to be cased, then returns to COSC for about five more days of testing as a complete watch.

The Excellence tolerance is -2/+4 seconds per day. The cased watch also goes through a wear simulation, a magnetism test at 200 gauss, and a check that it runs for its full marketed power reserve. A pilot started in March 2026, and the program is meant to be fully open to brands from October. Only watches a brand chooses to submit earn the label, so expect it to appear on some references rather than whole catalogs.

Because it's new, check the dial or papers rather than assuming. A standard COSC watch from a brand that's taken part in the pilot isn't automatically an Excellence watch.

METAS is the Swiss Federal Institute of Metrology. Its Master Chronometer certification was developed with Omega, and it tests the assembled watch, not the loose movement. The watch has to pass COSC-style testing first, then a further series of tests on the finished piece.

The accuracy window is far tighter than COSC: an average of 0 to +5 seconds per day, with sources noting up to +7 for smaller calibers. The watch is also exposed to a 15,000 gauss field (roughly MRI strength) and has to keep time afterward. Power reserve is checked against the brand's stated figure, accuracy is measured with the reserve partly run down, and water resistance is tested to the rated pressure plus a margin.

Omega is the brand most associated with METAS, and Tudor is increasingly pursuing it for selected models. Brands run the testing themselves under METAS audit, which is a difference from COSC's fully external lab.

Rolex builds on COSC. Movements are still sent for official certification, then Rolex tests the finished Oyster in-house to -2/+2 seconds per day after casing. Rolex also checks water resistance, self-winding performance and power reserve on the cased watch, and backs the result with a five-year guarantee.

On pure daily-rate tolerance, that's the tightest of the mainstream published standards, tighter than both COSC and METAS. The trade-off is that the testing is internal with no outside regulator, so you're trusting Rolex's documentation rather than a third-party certificate. Most owners reasonably do, and the guarantee gives that trust something to stand on.

Reports indicate Rolex tightened the standard in late 2015 and applied it across its production, which is why a modern Rolex tends to run closer to the limit than an older example of the same caliber.

Both run their own standards and don't rely on COSC as the headline. The Patek Philippe Seal sets -3/+2 seconds per day for movements over 20mm in diameter, with different limits for smaller calibers (-5/+4) and tourbillons (-2/+1). Testing runs through production and ends with an examination of the fully assembled watch.

The Grand Seiko Standard guarantees -3/+5 seconds per day for its mechanical watches, tested over 17 days across six positions and three temperatures. Grand Seiko also has a stricter Special Standard, marked on the dial. Its 9F quartz movements are specified at plus or minus 10 seconds per year, which is a different class of accuracy altogether.

Note that Grand Seiko's own manuals for individual 9S calibers quote a looser normal usage range for everyday wear. That's normal: a testing standard describes conditions in the lab, while a manual describes real use.

First, don't compare them as if they were identical tests. A -2/+2 figure on a cased watch and a -4/+6 figure on a bare movement measure different things. Cased and magnetically tested standards say more about how the watch behaves in daily life.

Second, a certificate describes the watch as it left the factory. Magnetization, shock, worn oil and an overdue service all pull a movement away from its tested rate. If a certified watch drifts well outside its standard, that's a reason to book a service and a demagnetization check, not a reason to doubt the standard.

When you're buying pre-owned, the certificate or papers tell you which standard applies, and a timegrapher reading from a watchmaker shows how close the watch still is. Ask for both.