Who Tested Your Watch? A Guide to the Watch Industry’s Quality Marks


People often wonder just how good their watches are. A quality seal can help provide an answer: it shows that the watch has passed a test for its timekeeping ability and, in some instances, the quality of its finishing. COSC and the Geneva Seal are by far the best-known issuers of quality certificates, but there are many others. Some, like these two, test watches from multiple brands; others, such as the Patek Philippe Seal. Here’s a look at 11 quality-testing systems.

COSC Certificate:

Switzerland’s Contrôle Officiel Suisse des Chronomètres, or COSC, is the best-known watch-certifying organization. The COSC procedure lasts 15 days. Watch companies send their movements to COSC uncased. COSC then assigns each an identification number and its employees apply a generic dial with reference points that help an optical sensor track the watch’s seconds hand. Each movement is wound mechanically (automatic movements must remain rotor-less), then given a rest period of 24 hours before testing. The optical sensor reads the time off the dial daily. This reading is logged to a server controlled by two atomic clocks, which provide a timing standard. During the first 10 days, the movement is tested in five positions, each for 48 hours: crown left, crown up, crown down, dial up and dial down. Movements are kept at a temperature of 23 degrees Cº, or 73.4 degrees Fº, during this period. If a movement has a chronograph function, it is activated on day 10 to check its effect on timekeeping. On days 11 through 13, the temperature is lowered to eight degrees Cº (46.4 degrees Fº), then returned to 23 degrees Cº, then raised to 38 degrees Cº (100.4 degrees Fº). For the last two days, the temperature is brought back down to 23 degrees Cº and the movement is returned to its original position with crown facing left.

At COSC, employees load uncased movements into trays before the 15-day test begins.
At COSC, employees load uncased movements into trays before the 15-day test begins.

COSC measures the movement’s performance against the seven criteria of ISO 3159, the international standard for what it calls “wrist-chronometers with spring balance oscillator.” (COSC has different criteria for movements with a diameter greater than 20 mm and those with a diameter of 20 mm or less. As most men’s watches have movements larger than 20 mm, all the criteria mentioned here are from the standard for that class.)

  1. Average daily rate. When the time is read off the dial each day, it is compared to an atomic standard. The movement’s daily deviation from atomic time is denoted in ±1/10s of a second. At the end of the test, these values are averaged to obtain an estimate of how fast or slow the movement generally runs. COSC takes the average of the first 10 days of the test (that is, before the chronograph, if it has one, is engaged, and before there is any variation in temperature). This number is the most commonly cited measure of a watch’s accuracy. For a watch movement larger than 20 mm in diameter, the average daily rate must fall between minus-four and plus-six seconds/day in order for the movement to be certified.
  1. Mean variation of rates. This test measures how consistent the rate is, even if it differs from the atomic standard. While average daily rate tells us how much a movement’s timekeeping deviates from an atomic standard, the variation tells us how much the movement deviates from itself, day to day. Since each test position (crown left, for example) is held for 48 hours, COSC measures the difference between daily rates on two consecutive days in the same position in order to know how much the timekeeping varies. For instance, if a movement runs 0.3 seconds slow on its first day with the dial up and 0.7 seconds slow on its second day in that position, then it has a variation of 0.4 seconds for those days. COSC averages the variation across the first 10 days of testing in order to get the mean variation of rates across all five positions. (This value is sometimes written as “Vmoy.”) COSC requires that the mean variation of rates be two seconds/day or less.

One way to think about average daily rate and mean variation of rates is as a matter of accuracy vs. precision. Accuracy is how close to perfect the results are – whether your dart hits a bull’s-eye, or how close your watch is to atomic time. Precision is a matter of consistency – if every shot hits the same spot on the board, even if it’s far from the bull’s-eye, that’s a sign of a very precise marksman. If a watch movement is precise, it keeps consistent time, and even if it’s not very accurate to the timing standard, it can be adjusted for better accuracy. So the mean variation is a good way to assess the quality of a movement.

  1. Maximum variation in rates. This is the greatest amount that the movement’s daily rate varies between the two days that it’s in any single position. The test is sometimes known as “Vmax.” COSC requires that the maximum variation be five seconds/day or less.
COSC issues a certificate for each movement, detailing its performance on each day of the test.
COSC issues a certificate for each movement, detailing its performance on each day of the test.
  1. Difference between rates in horizontal and vertical positions. Since gravity affects how a balance operates, there can be significant differences in a movement’s timekeeping when it is flat and when it is in a hanging position. For this calculation, COSC compares the average daily rate of days one and two (when the movement is vertical, with crown left) with the average daily rate of days nine and 10 (when the movement is horizontal, with the dial up). The COSC requirement is that the difference falls between minus-six and plus-eight seconds/day.
  1. Greatest deviation between the average daily rate and any individual rate. This is the largest difference between the average daily rate and the daily rate on any of the first 10 days. This calculation is sometimes referred to as the “largest variation in rates” or the “greatest spread of rates.” For COSC certification, it should be 10 seconds/day or less.
  1. Variation of rate per degree. To measure how much a change of temperature affects the movement’s timekeeping, COSC looks at the difference between the rates at eight and at 38 degrees C, then divides that rate by 30 to see how much timekeeping changes per each degree of temperature change. COSC requires the variation to be within ±0.6 seconds/day per degree.
  1. Rate resumption. In order to assess how much the various positions and temperatures have affected the movement cumulatively, COSC compares the movement’s rate at the end of the test to its rate at the start. The daily rate for the final day of testing, day 15, is compared to the average of the first two days’ rates. For COSC certification, this difference must be between minus-five and plus-five seconds.

Although COSC’s procedure is standardized and machine-operated to allow for testing of large batches, COSC also provides alternative testing methods. ISO 3159 only covers mechanical movements, but COSC has its own criteria for testing quartz watches. It tests them with the dial up for 11 days at three temperatures. The movements get rotated in three dimensions and must withstand 200 shocks of 100 Gs. COSC certified more than 46,000 quartz movements in 2012, the most recent year for which data is available. For an added fee, COSC can also test cased-up watches and perform a modified, manual version of the standard chronometer test. In 2012, COSC issued 1.73 million certificates; some 85 percent of those went to three brands: Rolex, Omega, and Breitling.

Geneva Seal:

The first requirement for the Geneva Seal, which dates back to 1886, is that the watch be made in the Canton of Geneva (both movement and watch must be assembled and given final inspection there) and that its manufacturer be based in the canton. The watch must also meet requirements for finishing and timekeeping and its complications must work correctly.  Certification begins with an examination of the movement to establish that its finishing meets Seal requirements. This occurs at the Seal’s testing facility, Timelab (the Geneva Laboratory of Horology and Microengineering).

Geneva Seal
Geneva Seal

The watch manufacturer submits to Timelab a reference set of all components, both external and internal, and an assembled movement. Officials there check the movement’s finishing, using a long list of criteria. Among them: bridges and plates must have polished chamfers and be decorated so as to remove machining marks; jewels must be set in polished holes and have polished sinks; wheels of the going train must be chamfered; the functional parts of the pivots and pivot shanks must be burnished; for movements larger than 18 mm in diameter, the escape wheel must be no more than 0.16 mm thick (or 0.13 for smaller movements); and the locking surfaces of the escape wheel teeth must be polished. If the testing lab approves the sample movement, the watch company uses it as a benchmark by which to measure the quality of all such movements it uses. Quality is monitored by the watch company itself, but Seal officials perform periodic audits to make sure that all movements produced match the reference movement.

Each cased-up movement is also tested, either at Timelab or on the manufacturer’s premises. Accuracy is measured over a period of seven days. The watch is placed on a machine that rotates it once a minute for 14 hours, then holds it in a single position for 10 hours. Chronographs are engaged for the first day of the test. In the course of the week, the watch must not lose more than a total of one minute. That comes out to an average daily rate of ±8.6 seconds/day. The watch’s additional functions – calendars, repeaters, equation of time indicators, etc. − are also tested. So is water resistance: all watches claiming to be water resistant must withstand pressure of 30 meters or higher, if a higher level is claimed for them. In addition, the watches’ power reserves must be at least as long as the manufacturers claim they are. Running time is measured while the watch is in the dial up position. The manufacturer is allowed to apply the Geneva Seal hallmark to the mainplate or to one of the bridges. Each Geneva Seal watch must be sold with its certificate, specifying the serial number of the movement and the case. Today, the Geneva Seal’s clients include Cartier, Chopard, Roger Dubuis (all of whose watches bear the seal), and Vacheron Constantin.

Fleurier Quality Foundation Seal:

To receive the Fleurier Quality Foundation seal, awarded by the foundation in the town of Fleurier in the Swiss Jura, a watch must meet five standards. It must be 100-percent manufactured in Switzerland; it must pass a Chronofiable test for durability (see next page); it must receive COSC certification; its movement’s finishing must pass a visual inspection, both by the naked eye and under a microsope; and it must undergo what the foundation calls the “Fleuritest.” If a watch meets all five requirements, the FQF logo is inscribed on its movement, and, if that inscription is not visible when the watch is cased up, on its dial as well.

This watch from Parmigiani Fleurier bears the FQF seal above the date aperture.
This watch from Parmigiani Fleurier bears the FQF seal above the date aperture.

In the Fleuritest, the cased-up watch is placed on a machine designed to mimic the typical motions of a human being. The machine moves the watch to simulate varying levels of activity, from almost zero (as when the wearer is asleep) to very high (as during vigorous exercise). The watch’s rate is measured using digital cameras that track the time on the dial. At the end of 24 hours, the rate must fall between zero and plus-five seconds. The seal was established in 2001. The participating companies are the watch brands Bovet, Chopard and Parmigiani and the movement-maker Manufacture Vaucher.

Besançon Observatory Certificate:

The Besançon  Observatory, in the former watchmaking center of Besançon, France, follows the international standards for chronometers provided by ISO 3159, and the testing procedure it uses is essentially identical to that of COSC. The major difference, however, is that Besançon tests only cased-up movements. That’s because the Observatory believes that adding complication plates and a dial and then casing up the movement can alter the watch’s rate, making a raw-movement test ineffective. When a watch passes the test, the watch company receives a rate certificate and can have the Besançon quality mark (a viper’s head) stamped on the mainplate. Besançon tests very few watches. It does not disclose who its clients are, but they include L. Leroy and Laurent Ferrier.

Glashütte Observatory Certificate:

The Glashütte Observatory, in the German watchmaking town of Glashütte, is an independent entity that operates in a facility that the German retailer and watch manufacturer Wempe purchased and refurbished in 2005. It is operated by the German Calibration Office and tests watches to meet German industrial standard DIN 8319, which is almost identical to ISO 3159. The observatory tests finished watches; it stipulates that they have a stop-seconds function so they can be set precisely to the second. The facility has its own test criteria for quartz watches. It does not reveal which brands use its certification system, but they include Wempe and Glashütte Original.

Wempe refurbished the original Glashütte Observatory, which was built between 1906 and 1910.
Wempe refurbished the original Glashütte Observatory, which was built between 1906 and 1910.

Chronofiable Certificate:

With a name based on the French word for “reliable” (fiable), the Chronofiable test has its origin in the 1970s, when the Federation of the Swiss Watch Industry established the Reliability Test Center (CCF SA) and made Chronofiable a trademarked testing procedure. Today Chronofiable is a department of Laboratoire Dubois SA, an independent horological testing center in La Chaux-de-Fonds. Chronofiable’s test parameters, according to the FH, remain unchanged after over 30 years. The premise of Chronofiable is to speed up normal watch wear by a factor of eight. Testing lasts 21 days, which Laboratoire Dubois claims is equivalent to six months of wear. This “aging cycle” includes a number of different elements. Watches are shock tested via linear accelerations that range between 25 and 550 Gs and angular accelerations of up to eight radians per second squared. The watch must also sustain temperature variations (15, 30 and 55 degrees Cº, or 59, 86, and 131 degrees Fº). Before and after this aging cycle, the watch’s timing, functions, and amplitude are checked. The laboratory tests representative sample movements, not each movement it certifies. Chronofiable testing is a requirement of the Fleurier Quality Seal (see above).

Patek Philippe Seal:

Established in 2009, the Patek Philippe Seal requires 19 days of testing and inspection. During this period, the cased-up movement receives a test of the power reserve, multiple measurements of its timing accuracy and amplitude, a kinetic simulation that mimics normal wear, a forced stopping of the balance to ensure it can restart itself under less than full wind, and inspections of sealing and water resistance. Additionally, there are numerous aesthetic inspections: requirements include satin-finishing for the sides of all steel parts, chamfering on the screw heads and decorations on the upper sides of bridges and on the winding rotor.

Patek Philippe seal requires 9 days of testing/inspection. Learn why Patek Philippe launched its own seal here.
Patek Philippe seal requires 9 days of testing/inspection. Learn why Patek Philippe launched its own seal here.

Timing accuracy is assessed at several points during production to ensure the accuracy of the raw movement as well as the final cased-up watch. For movements with a diameter of 20 mm or larger, the daily rate deviation must fall between minus-three and plus-two seconds/day. Each watch is tested in six positions. Every mechanical watch Patek Philippe produces must be tested and receive the Patek Philippe Seal before it is delivered to the consumer.

Grand Seiko Certificate:

The Grand Seiko Standard Inspection is used to test mechanical Grand Seiko movements before they are cased. The procedure is very similar to COSC’s. Dials and hands are applied to the movement and inspectors measure rate accuracy via photographs of the dial. Seiko has a stricter standard of accuracy than COSC: the average daily rate must be between minus-three and plus-five seconds/day. The mean variation of rate has to be 1.8 seconds/day or less, and the maximum variation between two days in one position is four seconds/day. The greatest difference between the mean daily rate and any individual day’s rate must be eight seconds/day or less.

Grand Seiko certifies all of its mechanical movements.
Grand Seiko certifies all of its mechanical movements.

Seiko tests its movements in six positions, rather than the five used by COSC. The brand argues that testing the movement with crown right (12 o’clock facing up) is an important addition, as a watch lying on a table and leaning back on its strap will be in this position. In order to accommodate this sixth position, Seiko inspects its watches for 17 days.

Jaeger-LeCoultre 1,000 Hours Control:

Jaeger-LeCoultre developed the 1,000 Hours Control in 1992. As the name suggests, it involves a battery of tests lasting nearly six weeks. Originally, the 1,000 Hours Control was used only for the in-house Caliber 889, but today every JLC watch leaving the factory (barring some unique pieces) must pass the inspection.

Machines Jaeger LeCoultre uses for its cyclotest
Machines Jaeger LeCoultre uses for its cyclotest

The procedure has six parts. In the first, the watch is placed in a machine that simulates the motions of daily wear to make sure all components are securely in place. Next, the amplitude of the balance wheel is tested in six positions. The third test is a verification of the power    reserve, ensuring that the movement runs as long as claimed. The effects of temperature are tested next: the watch begins at 22 degrees C (71.6 degrees F), and then is placed in environments at four and 40  degrees C (39.2 and 104 degrees F, respectively). The fifth part of the test is the longest: it’s the “cyclotest,” which averages two weeks in length. During it, the watch is alternately rotated and placed at rest for varying periods of time to test its timekeeping accuracy. (JLC does not disclose its tolerances.) Finally, the watch’s water resistance is tested at air pressures between three and 20 atmospheres, depending on the watch model. Watches that pass get the 1,000 Hours Control engraving on the caseback.

Ulysse Nardin Certificate:

Introduced in 2012, the Ulysse Nardin Chronometer & Performance Certificate has been used to test the in-house movements Caliber 118 (used in the Marine Chronometer Manufacture) and Caliber 150 (used in the Marine Chronograph Manufacture). The seven-day process includes an aesthetic inspection, a water-resistance test, and vacuum and pressure tests (simulating air pressure at -0.6 bar and +2.5 bar, respectively). But its main component is a five-day test of the watch’s accuracy. Each watch is tested in six positions (staying in each for 20 hours) and at three temperatures (23, eight and 38 degrees Cº). The range for the deviation of rate is narrower than COSC’s: minus-two to plus-six seconds/day. The deviation based on temperature is also slightly smaller, at ±0.5 seconds/day per degree.

The UN 118 is one of two manufacture movements that receive the Ulysse Nardin certificate.
The UN 118 is one of two manufacture movements that receive the Ulysse Nardin certificate.

Richard Mille Certificate:

Richard Mille introduced the RM 031 High Performance Certificate in 2012 for the RM 031 High Performance watch. In order to receive this guarantee of quality, the watch must have a rate deviation of less than 30 seconds a month, or one second/day. The watch is tested for 61 successive days. Initially, the movement is sent to COSC for a standard 15-day chronometry test. After it gets cased up at APRP (Audemars Piguet Renaud et Papi), it is put through a 15-day cyclotest, in which it rotates once a minute and timing is measured daily. If afterwards the rate deviation is more than 15 seconds (or one second/day), the watch is dismantled and assembly starts again from scratch. Finally, at Richard Mille’s factory in Les Breuleux, the cyclotest is repeated, now running 31 days. If the average monthly rate variation is less than 30 seconds, the watch gets a certificate listing the rate for each day.

Comparing the Rate Tolerances (Average Daily Rate Requirements):

Quality MarkAverage Daily Rate Tolerance (seconds/day)*

COSC Certificate:     -4 to +6

Geneva Seal:         -8.6 to +8.6

Fleurier Quality Foundation Seal:    0 to +5

Besançon Observatory Certificate:  -4 to +6

Glashütte Observatory Certificate:      -4 to +6

Patek Philippe Seal:                 -3 to +2

Grand Seiko Certificate:          -3 to +5

Ulysse Nardin Certificate:       -2 to +6

Richard Mille Certificate:        -1 to +1

* Average daily rate for movements larger than 20 mm

Some certifications do not differentiate based on size of movement.

No Responses to “Who Tested Your Watch? A Guide to the Watch Industry’s Quality Marks”

Show all responses
  1. Hugo Maiutto

    At last a very interesting and professional article.
    I´ll keep it in my records
    Thank you so much.

Leave a Reply