The Medieval City Under Issyk-Kul

Updated October 5, 2026 · 11 min read

The Medieval City Under Issyk-Kul
Original graphic by Kyrgyzstan Guides

Somewhere between one and four metres below the surface of Issyk-Kul, close enough to the shore that you could swim out to it, there's a medieval town. Fired-brick buildings. A mill with its millstone still in place. A Muslim cemetery covering about six hectares.

It was a working settlement on the Silk Road from roughly the tenth to the fifteenth century, and then an earthquake in the early 1400s put it under the lake.

Underwater surveys by archaeologists from the Russian Academy of Sciences have now mapped four zones of it, reported in detail by Heritage Daily. This is, by a distance, the most interesting thing in Kyrgyzstan that almost no visitor knows exists.

What has been found

The site corresponds to a settlement known as Toru-Aygyr, on the western part of the lake.

Why there are drowned sites at all is the lake-level story our post on Issyk-Kul’s lost outlet tells in full. Divers worked four separate zones in shallow water (nowhere deeper than about four metres) and found something different in each:

ZoneFinds
1Numerous fired-brick structures, one containing a millstone: a medieval flour mill
2A Muslim necropolis of the 13th–14th centuries, roughly 6 hectares
3Later expansion, more buildings, plus an earlier burial ground built over
4Round and rectangular mudbrick structures, with layers of buried soil

Read zone 3 again, because it's the archaeologically interesting one. An earlier cemetery with a later town built on top of it means the settlement grew, outlived its own memory of where its dead were, and expanded over them. That is a town with several centuries of continuous life behind it, not a seasonal camp.

The six-hectare necropolis in zone 2 is the size that tells you about population. Cemeteries scale with the communities that fill them, and that's a lot of graves.

Why a city is under a lake

Issyk-Kul is endorheic, nothing flows out of it. Its level is set entirely by the balance of inflow against evaporation, and that balance has shifted substantially over the centuries. The lake has been higher and lower than it is now, repeatedly.

On top of that slow movement sits the fast one: this is an extremely active seismic zone. The Tien Shan is still rising because the Indian plate is still pushing north, and the faults that do that run directly under and around the lake. Our guide to earthquakes in Kyrgyzstan covers the mechanics.

A large earthquake can drop a section of shoreline, liquefy the ground beneath buildings, or trigger landslides that change a river's course. Researchers connect the loss of this settlement to a powerful event in the early fifteenth century; Live Science's account frames it as a Silk Road hub wiped out by a single earthquake.

The result is a specific and rare kind of archaeological site: not a ruin that decayed slowly with people picking over it for eight hundred years, but a town that went under in a short space of time and has been sitting in cold, low-oxygen fresh water ever since.

This isn't the only one

Issyk-Kul has produced submerged material for well over a century, and local reports of walls visible in clear shallow water go back much further.

Various settlements around the shoreline have yielded pottery, coins, bricks, millstones and human remains in shallow water, and the lake has a long association in Kyrgyz oral tradition with the idea of a drowned city. The tradition, it turns out, wasn't metaphorical.

This should adjust how you look at the lake. Issyk-Kul is presented to visitors as scenery — a beach, a backdrop, a place to swim between mountain trips. Our Issyk-Kul guide treats it that way too, because that's what most people come for. But the shoreline you're sitting on isn't where the shoreline was, and the shallow water in front of you is an archaeological deposit.

Can you see any of it?

Manage expectations before you plan anything around this.

There is no visitor site. No dive centre running trips to the sunken city, no marked underwater trail, no interpretive centre. The survey work is research, conducted by permit.

Some material is in museums. The Issyk-Kul regional museum at Cholpon-Ata and museums in Bishkek hold finds recovered from the lake over the decades, pottery, bronze items, coins. This is the realistic way to see it.

Do not go looking yourself. This matters. Removing anything from an archaeological site is theft of cultural property, the law here treats it seriously, and shallow-water sites are exactly the ones that get destroyed by amateur collecting. If you snorkel and you see worked stone or brick, leave it where it is.

Diving in the lake is possible through operators on the north shore, but as recreational diving in a high-altitude lake rather than as archaeology. Altitude diving has its own decompression rules — Issyk-Kul is at 1,600 m, and standard dive tables don't apply without correction. DAN's guidance on altitude diving is the thing to read before you book, not after.

Why it's worth knowing

Because it reframes the country you are travelling through.

Kyrgyzstan is sold as a nomadic culture, and that's true and central. It also had towns: caravan towns on a working trade network, with mills and brick buildings and cemeteries large enough to imply real populations, in the same valleys now used for summer pasture. The settled layer of this country's history is genuinely under-told, partly because so much of it's under a lake, under silt, or was dismantled for building material centuries ago.

The other thing it tells you is that this landscape moves. The Black Death cemeteries of the Chuy valley, this town, the earthquake-shattered sites elsewhere: the mountains that make Kyrgyzstan beautiful are the same mechanism that repeatedly erased what people built in them.

How much the lake actually moves, and what is really dated

The submerged sites make sense only once you know how far this lake rises and falls, and the measured record is dramatic enough without embellishment.

Instrumental records begin in 1926. Between 1927 and 1997 the level fell 3.4 metres; from 1997 to 2011 it rose 0.93 m. But the historical swing is far larger. Work on dated shorelines indicates the lake probably stood below 1600 m in the late fifteenth century, then rose to about 1619 m over thirty to forty years — leaving a wave-cut scarp some 10 m high — before falling roughly 8 m over the following century. The Chu River outflow caps it at about 1620 m. Seismic profiling has identified at least four stepwise cycles of 400 m or more over geological time.

There is a second mechanism too, and it is not about water at all. A paleoseismologist concluded that a strong earthquake between the fourteenth and sixteenth centuries dropped the northern shore below the waterline. No written source records that event.

The survey history is longer than most accounts suggest. Diving seasons began under D. F. Vinnik in 1958. From 1984 the Kyrgyz Academy of Sciences expedition under V. M. Ploskikh worked the sites, publishing Issyk-Kul: sunken cities in 1988, and after 1993 a joint expedition with the Kyrgyz-Russian Slavic University ran regular seasons to about 2013. The main site at Toru-Aygyr stretches 5 to 6 km along the shore and up to 0.7 km offshore, at depths of 0.5 to 5 m.

Here is the part to be careful about. The dating is far thinner than the coverage implies. The first and only radiocarbon date from underwater came from a foundation pile at Sary-Oy during 2012–14 work, confirming an early fifteenth-century date — and it was never published. Everything else is dated by ceramic and brick typology. Trees found in situ in 2025 are still awaiting dating.

Which brings us to the claim you will meet everywhere: a 2,500-year-old sunken civilisation. It traces to a single news agency article, apparently from December 2007, reporting stone walls of about half a kilometre at 5 to 10 m depth, wave-eroded Saka kurgans, bronze axes and arrowheads, akinaki daggers and a 70 g gold pin. No dating method is stated anywhere in it — the age is an inference from artefact style. The peer-reviewed 2024 survey of the same waters describes a kurgan, a necropolis and a medieval settlement, not a Saka city. Saka-period material under this lake is real; a 2,500-year-old sunken urban civilisation is not established.

FAQ

Is there a sunken city in Lake Issyk-Kul?

Yes. Underwater surveys by the Russian Academy of Sciences have mapped a medieval settlement at depths of one to four metres, including fired-brick buildings, a mill and a large cemetery, corresponding to the site of Toru-Aygyr.

How old is the sunken city at Issyk-Kul?

The settlement flourished from around the tenth to the fifteenth century, with the necropolis dated to the thirteenth and fourteenth centuries.

Why did the city sink?

Researchers link its submergence to a powerful earthquake in the early 1400s. Issyk-Kul sits in a highly active seismic zone and its water level has also varied considerably over the centuries.

How deep is the sunken city?

Between about one and four metres, shallow enough for divers to work in, and in places shallow enough to see into from the surface in calm, clear conditions.

Can tourists dive the sunken city?

Not as an organised archaeological experience. There is no visitor site or dedicated dive trip. Recreational diving is available on the lake, but this is a research site and removing objects is illegal.

Where can I see finds from Lake Issyk-Kul?

The regional museum at Cholpon-Ata and museums in Bishkek display pottery, coins and metalwork recovered from the lake over many decades.

How much does the level of Issyk-Kul change?

A great deal. Instrumental records from 1926 show a 3.4 metre fall between 1927 and 1997 and a 0.93 metre rise to 2011. Historically the swing was larger: the lake probably stood below 1600 metres in the late fifteenth century before rising to about 1619 metres within thirty to forty years, leaving a wave-cut scarp around ten metres high.

Is there really a 2,500-year-old city under Issyk-Kul?

Saka-period material is genuinely reported, including wave-eroded kurgans, bronze weapons and a gold pin, but the 2,500-year figure traces to a single 2007 news article that states no dating method — the age is inferred from artefact style. The peer-reviewed 2024 survey describes a kurgan, a necropolis and a medieval settlement rather than a Saka city.

Has anything under Issyk-Kul been radiocarbon dated?

Only once, and it was never published. A foundation pile at Sary-Oy was dated during work in 2012-14, confirming an early fifteenth-century date. Everything else rests on ceramic and brick typology, and timbers found in situ in 2025 are still awaiting dating.

Why are there settlements under the lake?

Two mechanisms, not one. The lake's level has risen and fallen by many metres over the historical period, and separately a paleoseismologist concluded that a strong earthquake between the fourteenth and sixteenth centuries dropped the northern shore below the waterline. No written source records that earthquake.

Toofan Singh
Written by
Toofan Singh

Toofan Singh is an India-based traveler and the founder of Kyrgyzstan Guides. He built the site as a research-led resource for trip planners: every guide is compiled from official sources, current operator prices and recent traveler reports, then updated whenever visa rules, transport costs or trail conditions change.