When Was the Antikythera Mechanism Made? — The Evidence for Its Date
The Antikythera mechanism was made in the 2nd or early 1st century BC. It must be older than the shipwreck it was found in, which sank around 70–60 BC. Proposed construction dates range from about 205 BC to around 87 BC, and researchers distinguish the date it was built from the “start date” set on its dials, which some place at 205, 204 or 178 BC.
A date on the mechanism’s dials is not the same as the date it was made. Its eclipse and calendar dials had to start somewhere, and that starting point could be years before or after the device was actually built.
The Antikythera mechanism was made by Greek craftsmen sometime in the second century BC or the early first century BC. Nobody has found a maker’s signature or a dated inscription, so its age has to be worked out from several independent clues: the date of the ship that carried it, the coins and pottery found with it, the style of the Greek letters engraved on it, the astronomy it was built to calculate and the starting points of its calendar and eclipse dials. Together they point to a window of roughly 205 to 60 BC, with most scholars favouring the later second century BC.
That makes the device more than 2,000 years old. Nothing of comparable mechanical complexity survives from anywhere in the world until the astronomical clocks of fourteenth-century Europe, more than a thousand years later, which is why its date matters so much: it is the clearest evidence we have that Hellenistic Greek engineers could build precision gearing long before the medieval clockmakers.
How does the shipwreck date the mechanism?
It sets the latest possible date. The mechanism was recovered in 1901 from the wreck of a large cargo ship that sank off Point Glyphadia on the small Greek island of Antikythera, between Kythera and Crete, at a depth of about 45 metres. Sponge divers from the island of Symi, led by Captain Dimitrios Kontos, found the wreck in 1900, and the Greek government organised one of the first major underwater archaeological salvage operations, with help from the Hellenic Royal Navy, to recover its cargo of bronze and marble statues, glassware, jewellery, pottery and coins. The divers worked in heavy standard dress at a depth that pushed the limits of the equipment of the day.
The ship itself has been dated by several lines of evidence to roughly 70–60 BC. The pottery on board, the coins and the style of the other artworks all fit that period. Since the mechanism was already on the ship when it sank, it had to have been built before then. How long before is the harder question. The device was not new when it went down: it was a finished, working instrument, and it may have been years or even decades old.

What do the coins tell us?
Coins are among the most useful dating tools in archaeology, because they carry the marks of the cities and authorities that minted them. Coins recovered from the wreck, including those found during the Jacques Cousteau expedition in the 1970s, include issues from Pergamon in Asia Minor and from other eastern Mediterranean cities. They help pin down when the ship sailed and suggest where its cargo came from.
Some researchers have gone further and argued that the coins point to the device itself having been made around the time they were struck, possibly in Pergamon, a wealthy Hellenistic city with a great library second only to Alexandria’s. That link is not proven: a ship’s purse tells you about its last voyage, not necessarily about the origins of every object on board. But the coins do confirm that the wreck belongs to the first half of the first century BC.
What do the inscriptions reveal about its age?
The mechanism was covered in engraved Greek text: labels on its dials, and long explanatory passages on its front and back plates that served as a kind of user’s manual. Thousands of characters survive, many readable only through X-ray computed tomography carried out from 2005 onwards by the Antikythera Mechanism Research Project.
Greek letter shapes changed over the centuries, and epigraphers, the specialists in ancient inscriptions, can date a text by the style of its lettering. The forms on the mechanism are typical of the second or early first century BC. Derek de Solla Price, who studied the fragments from the 1950s and published his landmark analysis in 1974, favoured a date in the first century BC; later work on the letter forms has generally pushed the estimate back into the second half of the second century BC.
What does the astronomy say?
The mechanism is a machine built to embody astronomical theories, so it cannot be older than those theories. Its most sophisticated feature is the way it models the Moon’s varying speed: the Moon moves faster across the sky when it is nearer the Earth and slower when it is farther away. The mechanism reproduces this with a clever pin-and-slot arrangement of two gears on slightly offset axes. This lunar theory is associated with Hipparchus of Rhodes, the great astronomer who was active from about 140 to 120 BC.
If the device uses Hipparchus’s theory, it was probably built during or after his working life, which points to the late second century BC. Some researchers have argued that the mechanism instead relies on earlier, Babylonian-style arithmetic methods for predicting eclipses, which would allow an earlier date. That debate is part of why the range of proposed construction dates is so wide.
What are the “start dates” on the dials?
The back of the mechanism carries two great spiral dials. The upper one, the Metonic dial, tracks a 19-year cycle of 235 lunar months used to keep lunar and solar calendars in step. The lower one, the Saros dial, covers 223 lunar months, about 18 years, after which eclipses repeat in a similar pattern. Each dial had to begin at a particular moment, and the predictions engraved in its cells reflect that starting point.
In 2014, researchers Christián Carman and James Evans studied the pattern of eclipse predictions on the Saros dial and argued that the dial’s cycle began shortly after the new moon of 28 April 205 BC. On that view the mechanism, or at least its design, could date to around 200 BC. In 2022, another team proposed that the mechanism’s initial calibration date was 23 December 178 BC, while other experts have argued for 204 BC.
These start dates do not necessarily tell us when the bronze was cut. An astronomical instrument can be set to begin at a convenient past epoch, just as a modern calendar program can display dates long before it was written. A dial starting in 205 BC could have been built decades later, using tables that began then. That is why many scholars now speak of a design or epoch date on the one hand and a construction date on the other.
Who might have built it, and does that help date it?
Partly. The month names on the Metonic dial are of the Corinthian type and closely match those used in Epirus in north-western Greece and on Corfu, ancient Corcyra. The Games dial, which tracks the four-year cycle of the Panhellenic athletic festivals, includes not only the famous games at Olympia, Delphi, Nemea and Isthmia but also smaller festivals: the Naa at Dodona in Epirus and the Halieia of Rhodes.
Rhodes was a centre of astronomy and mechanical engineering in the second and first centuries BC, home to Hipparchus for part of his career and later to the Stoic philosopher Posidonius, who is said to have built a similar instrument. Some scholars have suggested the device came from a Rhodian workshop. A 2017 study by Paul Iversen argued that the prototype was Rhodian, but that this particular example was adapted for a client in Epirus, and that it was probably built no earlier than about a generation before the shipwreck. Alexander Jones, one of the leading scholars of the mechanism, has also supported a date relatively close to the wreck.
So which date is most likely?
There is no single agreed answer, but the evidence clusters. A construction date between about 150 and 100 BC is the range most often cited, and it fits the letter forms, the use of Hipparchus’s lunar theory and the date of the shipwreck. Earlier dates around 200 BC rest mainly on interpreting the Saros dial start date as a construction date, while a date around 87 BC reflects older readings of the inscriptions and coins.
What every serious proposal agrees on is the big picture: the mechanism was built in the Hellenistic world, more than 2,000 years ago, by people with an advanced understanding of astronomy and an astonishing level of metalworking skill. Whichever end of the range is right, it predates comparable geared machines by well over a millennium. The disagreements are about decades, not centuries, and they come from reading the same evidence in different ways rather than from any doubt that the device is genuinely ancient.
Why are there no similar devices from the same period?
Almost certainly because bronze was valuable and was melted down. The mechanism’s quality suggests a tradition of instrument-making, not a one-off invention: the gearing is too refined to have been a first attempt. Ancient writers describe similar devices. Cicero, writing in the 50s BC, mentions two celestial globes or planetaria built by Archimedes, which the Roman general Marcellus brought to Rome after the sack of Syracuse in 212 BC. Cicero also refers to a device made by Posidonius that showed the motions of the Sun, Moon and planets.
None of those instruments survives. The Antikythera mechanism did so only because the ship carrying it sank, and because the sea preserved, in corroded form, what would otherwise have been recycled. Its survival is a reminder that the archaeological record of ancient technology is a very small sample of what once existed.
How was its age recognised after it was found?
Slowly. The mechanism came up from the wreck in 1901 as a lump of corroded bronze and wood and was stored with the other finds at the National Archaeological Museum in Athens. For about two years nobody paid it much attention, while museum staff concentrated on the statues. On 17 May 1902, archaeologist Valerios Stais and his cousin, the politician Spyridon Stais, noticed a gear wheel embedded in one of the pieces.
At first many scholars refused to believe it was ancient. A geared astronomical device seemed too advanced for the period of the other finds, and some suggested it was a later object that had fallen onto the wreck. Detailed study by Albert Rehm, then by Price from 1951, and finally by the Antikythera Mechanism Research Project using X-ray tomography, showed beyond doubt that it belonged with the rest of the cargo and was as old as the ship’s other contents.
Could new finds settle the date?
Possibly. Divers have returned to the Antikythera wreck repeatedly, including a series of investigations that ran from 2014 to 2019 and a new five-year programme that began in 2020. They have recovered more artworks, evidence of a possible second ship nearby and a bronze disc decorated with a bull that was briefly suspected of being part of the mechanism, though it was more likely a furniture fitting. Another fragment, known as Fragment F, was found in 2005 in museum storage, having gone unnoticed since the original salvage.
Any new piece carrying inscriptions or gearing could add evidence for the date. So could fresh scientific analysis of the fragments themselves. Over the past two decades, high-resolution imaging has transformed what can be read on the surviving pieces, and each new reading of the inscriptions has fed back into the dating debate. In 2021 a team at University College London published a reconstruction of the whole device, including the planetary display, and the most recent studies continue to test how accurately it could have worked.
For now, the safest summary is the one most specialists would give: made in the Hellenistic period, most probably in the second half of the second century BC, and certainly before the ship sank in the first half of the first century BC.
How old is the Antikythera mechanism?
More than 2,000 years old. It was built in the 2nd or early 1st century BC.
When did the Antikythera ship sink?
Around 70–60 BC, based on its coins, pottery and other cargo.
When was the Antikythera mechanism discovered?
It was recovered from the wreck in 1901; a gear was first noticed in the fragments on 17 May 1902.
Who made the Antikythera mechanism?
Unknown. Hellenistic Greek craftsmen, possibly linked to Rhodes or Corinthian colonies; Hipparchus’s astronomy may have influenced it.
Is the Antikythera mechanism older than Archimedes?
Probably not. Archimedes died in 212 BC; most proposed dates are later, though some designs may draw on his tradition.
What does the 205 BC date refer to?
The proposed start date of the Saros eclipse dial, not necessarily the year it was built.
When did similar machines appear again?
In the 14th century, with the astronomical clocks of Richard of Wallingford and Giovanni de’ Dondi.
- Freeth, T. et al. (2006, 2008), Nature: decoding the Antikythera mechanism; calendars and the Games dial
- Carman, C. & Evans, J. (2014), on the epoch of the Antikythera mechanism and its eclipse predictor
- Iversen, P. (2017), the calendar on the Antikythera mechanism and the Corinthian family of calendars
- Price, D. de S. (1974), Gears from the Greeks
Blane Perun has been exploring Earth’s oceans and marine life for over 25 years. Founder of TheSea.Org in 1999, underwater photographer, coral aquaculture pioneer, and explorer of 80+ countries.
About Blane Perun →