In 1963, a man in the central Anatolian town of Derinkuyu knocked through a wall in his basement and found a room he didn’t know existed. Behind that room was another. And another. And then a passageway dropping into darkness that, as Turkish authorities would eventually confirm, descended eighteen stories — roughly 85 meters — into the volcanic tuff beneath his house. The popular version of the story says his chickens kept vanishing into a crack in the floor. Whether or not poultry were involved, what emerged from behind that wall was one of the most disorienting archaeological revelations of the twentieth century: an entire city, built for an estimated 20,000 people, hidden beneath a small town where residents had been casually living on top of it for generations.

Derinkuyu is now the most famous underground city in Cappadocia. But here’s the thing almost nobody talks about: it’s not alone. Not by a long shot.

The Number Nobody Can Agree On

How many underground cities exist in Cappadocia? The answer depends on who you ask and what you mean by “city.”

Between 1991 and 2000, the Italian research group Centro Studi Sotterranei — led by Roberto Bixio out of Genoa — conducted the most systematic survey anyone has attempted. Across six districts of historical Cappadocia, they documented 183 underground sites. Their findings were published as Cappadocia: Records of the Underground Sites through BAR International Series, and the dataset remains the most comprehensive catalog available. Other estimates push the count above 200 when you include sites across the broader region that haven’t been formally surveyed.

But this number is misleading if you picture 200 Derinkuyus. Most of these sites are small — single-level refuges, storage chambers carved into soft tuff, shelters that might have hidden a few families during a raid. The multi-story, infrastructure-heavy complexes with ventilation shafts, wells, rolling stone doors, and communal spaces number far fewer. Perhaps 40 have been formally identified by Turkish authorities. Maybe six to eight are what any honest archaeologist would call a “city.”

The gap between “200 underground sites” and “6 excavated cities open to tourists” is where the real story lives.

What’s Actually Open — and What Isn’t

If you visit Cappadocia today, you can descend into Derinkuyu (eight of its eighteen floors are accessible), Kaymakli (four of an estimated eight floors), Ozkonak, Tatlarin, Mazi, and Gaziemir. In 2020, the Turkish government opened Kayaşehir in Nevşehir, which was discovered in 2013 when a government housing project broke through into cavities during construction. Initial reports called it the “world’s largest underground city,” estimating its area at 460,000 square meters — roughly 30 percent larger than Derinkuyu. Over 15,000 air ducts were found. Murat Gülyaz, director of the Nevşehir Museum and lead archaeologist on the site, publicly cautioned against the superlative, noting that the site is structurally different from the classic tunnel-connected cities. Kayaşehir is more accurately described as a hillside cliff dwelling complex. The rooms aren’t interconnected through continuous tunnel systems the way Derinkuyu and Kaymakli are.

So that gives you maybe seven or eight sites you can visit. What about the other 175?

Some are on private land. Some are partially collapsed. Some are known only from Bixio’s surveys and have never been formally excavated. And some sit beneath modern towns where people live and work, which creates a problem that no amount of archaeological enthusiasm can easily solve: you can’t dig up someone’s house because there might be a Byzantine tunnel underneath it.

Turkey’s Law No. 2863 on the Conservation of Cultural and Natural Property is the statute that governs all of this. Its provisions are blunt: all cultural property, above and below ground, belongs to the state — whether it has been discovered yet or not. Excavation requires a formal permit from the Ministry of Culture and Tourism. Applications demand detailed project plans, documentation of research objectives, and for foreign teams, navigation of a bureaucratic process that the American Research Institute in Turkey (ARIT) helps facilitate.

Accidental discoveries must be immediately reported. Unauthorized excavation carries serious penalties. The intent is preservation, and on paper, the framework is reasonable. Cappadocia also falls within the boundaries of a UNESCO World Heritage Site — “Göreme National Park and the Rock Sites of Cappadocia,” inscribed in 1985 — which adds another layer of international oversight.

But the practical effect of this legal architecture is a kind of enforced stasis. With limited budgets, a small number of qualified specialists, and hundreds of known sites competing for attention, the Ministry has little incentive to open new excavations when it can barely maintain the ones already attracting tourists. Derinkuyu alone requires ongoing structural monitoring, ventilation management, and crowd control. Why take on another site when the existing ones are already straining resources?

There’s also a less discussed dynamic: the tension between heritage protection and development. Al-Monitor has reported on construction pressure encroaching on Cappadocia’s heritage sites, with critics accusing the ruling party of favoring concrete over conservation. The Kayaşehir discovery is itself a product of this tension — a housing project literally broke through into an underground city. The government pivoted from demolition to tourism, but the incident raised an uncomfortable question: how many sites have been damaged or destroyed by construction projects that didn’t happen to punch through into something photogenic enough to save?

Who Built Them, and Why?

The honest answer is: we’re not entirely sure, and the timeline is contested.

The oldest construction phases may date to the Hittites around 2000 BC or the Phrygians around 700 BC, but attributing specific tunnels to specific civilizations is difficult when the same spaces were modified, expanded, and repurposed over millennia. What the archaeological evidence does support is that the cities reached their fullest development during the Byzantine era, roughly the 5th through 10th centuries. Most datable artifacts recovered from sites like Derinkuyu and Kaymakli correspond to the Middle Byzantine Period.

The functional logic is clearer than the timeline. Cappadocia sits on a high Anatolian plateau that has been, for most of recorded history, a corridor for invading armies. Arab-Byzantine wars, Mongol incursions, Ottoman expansion — the region was repeatedly overrun. The underground cities weren’t permanent residences. They were refuges. The rolling stone doors (massive circular stones that could be rolled across tunnel entrances from the inside but not from the outside), the narrow corridors designed so that only one person could pass at a time, the ventilation shafts that could be sealed, the wells that provided water independent of the surface — all of it points to a population that expected to be besieged and engineered accordingly.

R.M. Dawkins, a Cambridge linguist who studied Cappadocian Greek-speaking communities between 1909 and 1911, documented that knowledge of the underground refuges was still alive in local oral tradition. The Greek Christian communities of Cappadocia used these spaces into the early twentieth century, before the population exchanges of 1923 severed the living chain of knowledge about many sites.

That severance matters. When the Greek-speaking population left, they took with them centuries of accumulated knowledge about which tunnels led where, which chambers were safe, which had collapsed. The Turkish communities that remained or resettled in the area often had no inherited understanding of what lay beneath their feet. Derinkuyu’s rediscovery in 1963 was, in a real sense, the rediscovery of a world that an entire displaced population had known intimately but could no longer access.

The Surveying Problem

Mapping what you can’t see is expensive, slow, and technically demanding. Geophysicists from Nevşehir University have used electrical resistivity tomography and seismic surveys to detect underground voids without excavation, but these methods produce probabilistic maps, not blueprints. They can tell you something is down there. They can’t tell you whether it’s a Byzantine chapel or a natural cavity in the tuff.

Bixio’s team from Centro Studi Sotterranei — working with collaborators Vittorio Castellani and Andrea de Pascale — did the most they could with surface surveys, entrance documentation, and limited subsurface exploration. But their project was fundamentally a cataloging exercise, not an excavation program. They recorded what could be accessed through existing openings. The sealed passages, the collapsed tunnels, the entrances hidden beneath modern construction — those remain uncharted.

The OBRUK Cave Research Group conducted a separate Underground Structures Inventory Project in Kayseri province, systematically surveying 33 underground cities in that area alone. Their work reinforced a pattern that Bixio’s surveys had already suggested: for every site that’s been formally studied, there are several more that exist only as entries in a database, awaiting funding and permits that may never come.

More recently, Sezin Nas at Istanbul Galata University has been studying the acoustics and soundscapes of the underground cities — a research angle that tells us something about how these spaces were actually experienced by the people who lived in them. Sound travels strangely underground. The echo patterns in Derinkuyu’s ventilation shafts would have carried voices, warnings, prayers through rock in ways we can only approximate now.

What Gets Restricted, and Why It Matters

The Turkish government doesn’t publish a list of sites it has decided not to excavate, and the reasons for inaction vary. Some are genuinely structural — a site is too unstable to enter safely. Some are jurisdictional — a site sits beneath private property or modern infrastructure. Some are financial — excavation costs money that the Ministry would rather spend on tourism infrastructure for existing sites.

But there’s also a category that’s harder to pin down: sites where excavation might complicate preferred historical narratives. Cappadocia’s underground cities were built and used primarily by Greek-speaking Christian populations. The churches carved into the rock, the crosses etched into tunnel walls, the Byzantine frescoes at Tatlarin — these are artifacts of a civilization that modern Turkey has a complicated relationship with. This isn’t a conspiracy theory. It’s a well-documented tension in Turkish heritage policy, one that scholars like those published in Internet Archaeology have analyzed in detail. The state protects these sites, but it doesn’t always rush to foreground their specifically Greek Christian character.

None of this means the government is hiding anything dramatic. The more mundane explanation — insufficient funding, bureaucratic inertia, legitimate safety concerns — accounts for most of the inaction. But the effect is the same regardless of the cause: the vast majority of Cappadocia’s underground world remains unstudied, unmapped, and inaccessible to both researchers and the public.

The Scale of What We Don’t Know

Consider the reported tunnel connecting Derinkuyu to Kaymakli — a passage said to stretch roughly nine kilometers underground. If it exists in its entirety (portions have been confirmed, the full length has not), it would represent an engineering project of staggering ambition, carved by hand through volcanic rock. Who organized that labor? How long did it take? Was it built all at once or extended over centuries? We don’t know. The tunnel isn’t fully excavated. It may never be.

Or consider the simple arithmetic: Derinkuyu’s eight accessible floors represent less than half of the estimated total. Below the tourist paths, below the gift shop and the crowd barriers, there are ten more levels that no one alive has systematically explored. What’s down there? Probably more of the same — storage rooms, living quarters, possibly cisterns. But “probably” is doing a lot of work in that sentence.

In 2025 and 2026, rescue excavations at the Göreme Open-Air Museum have been uncovering new underground settlement traces. Authorities also caught two individuals conducting illegal excavation at a previously unregistered underground city — a reminder that the demand to know what’s down there extends well beyond the academic community. Looters have their own theories about what the sealed tunnels contain.

The question that nags is not whether there are more underground cities in Cappadocia. There are. We know this. The question is whether we will ever develop the political will, the funding, and the institutional capacity to study them before they collapse, get built over, or are quietly looted into oblivion. Bixio’s surveys are now over two decades old. The researchers who conducted them are aging out of the field. No comparable project has replaced theirs.

Somewhere under a house in Nevşehir province, behind a basement wall that nobody has knocked through yet, there is almost certainly another room. And behind that room, another. And we are running out of reasons not to look.