The Toba Catastrophe: The Supervolcano That Nearly Wiped Out Humanity 74,000 Years Ago

About 74,000 years ago, a supervolcano in what is now northern Sumatra, Indonesia, produced the largest eruption in the past 28 million years. An estimated 1,700 cubic miles of rock, ash, and gas blasted into the atmosphere. That is a volume comparable to three million Empire State Buildings worth of material. The crater that formed afterward is now Lake Toba, roughly 60 miles long and 20 miles wide, one of the largest volcanic lakes on Earth. Ash from the eruption settled as far as East Africa, 5,000 miles away.

For a period in the 1990s and 2000s, scientists thought this eruption had nearly wiped out our species. The hypothesis was compelling, the evidence seemed to fit, and it spread widely. The story of the Toba catastrophe is still one of the most dramatic in human prehistory. It is also, increasingly, a story about how science corrects itself.

The Scale of What Happened

Even setting aside the debate about its human impact, the Toba eruption was genuinely extraordinary. A VEI 8 event, the highest rating on the volcanic explosivity index, it released sulfur dioxide into the stratosphere in quantities that filtered sunlight globally. Ice cores from Greenland show a significant sulfate spike at 74,000 years ago, followed by evidence of cooling. Some models suggest global temperatures dropped by 3 to 5 degrees Fahrenheit on average, with some regions seeing far sharper declines. The eruption likely caused a volcanic winter lasting several years.

Parts of the Indian subcontinent received 15 centimeters of volcanic ash. Parts of Indonesia were buried far deeper. The immediate environmental impact in South and Southeast Asia was severe. Plants cannot grow under centimeters of ash. Animals that depend on plants cannot eat. The food chain takes time to recover.

The ash layer from Toba is now a useful stratigraphic marker. Geologists find it in sediment cores across the Indian Ocean, the Arabian Sea, and lake beds in Africa. It is one of the clearest physical records of any single event in the Pleistocene.

The Bottleneck Hypothesis

In 1998, anthropologist Stanley Ambrose published a paper connecting Toba to something geneticists had been noticing in human DNA. Modern humans have unusually low genetic diversity compared to other great apes. Chimpanzees, despite having a smaller population than humans today, are genetically more diverse than we are. This pattern in human genetics is consistent with what biologists call a bottleneck: an event that dramatically reduced the human population, eliminating most existing genetic variation. Only the descendants of the survivors carry genetic information today.

Ambrose proposed that Toba caused this bottleneck. The volcanic winter, he argued, killed most humans everywhere except possibly in Africa, where equatorial conditions may have buffered the worst effects. Human population dropped to somewhere between 1,000 and 10,000 individuals. All of us alive today descend from that tiny remnant.

The hypothesis was compelling partly because of timing. Genetic analysis had placed the human population bottleneck somewhere between 50,000 and 100,000 years ago. Toba at 74,000 years fell within that window. The connection seemed to fit.

Why the Evidence Got More Complicated

Here is where it gets interesting, because the Toba catastrophe hypothesis has taken significant hits from multiple directions. Archaeological evidence from India is the most direct problem. Sites both above and below the Toba ash layer, meaning sites from before and after the eruption, show continuous human occupation with similar tools and similar behaviors. The people were there before the eruption. The people were there after. The ash layer sits between the two, but the human activity on either side looks continuous rather than interrupted by a near-extinction event.

Similar evidence has accumulated from South Africa and Arabia. Wherever archaeologists have looked carefully at sites spanning the Toba eruption period, they tend to find humans persisting rather than disappearing. A 2026 study published in Science Daily summarized the growing consensus: “As the evidence accumulates, it appears that people were able to survive and continue to be productive after Toba blew its stack.”

Climate modeling has also weakened the hypothesis. More recent simulations, accounting for feedback effects that earlier models missed, suggest the volcanic winter may have been less severe than originally calculated, particularly in Africa. Vegetation records from East African lake sediments show limited disruption around 74,000 years ago.

The genetic bottleneck itself is now also contested. Recent analysis suggests the apparent low genetic diversity in modern humans may have been building for much longer than Toba’s timeline, possibly related to population dynamics 100,000 to 150,000 years ago, before the eruption. The timing does not line up as cleanly as it first appeared.

Most working paleoanthropologists now consider the strong version of the Toba catastrophe hypothesis, that the eruption nearly caused human extinction, to be unlikely. What replaced it is more nuanced.

What Researchers Think Now

The current view is roughly this: Toba was a massive eruption with real environmental consequences. It almost certainly caused significant disruption and local or regional population declines. Human groups in some areas probably suffered badly. But the eruption did not reduce our species to a few thousand survivors on the edge of extinction. Humans in various locations appear to have weathered it, some more successfully than others.

The human population bottleneck that shows up in genetics was real. Its cause is still debated. Toba may have contributed. Climate changes unrelated to Toba, disease, or other factors may have been equally or more important. The bottleneck may also be an artifact of how geneticists model ancient population dynamics, with assumptions that are being revised as ancient DNA techniques improve.

What makes this story worth telling even after the dramatic original version has been largely revised is what it reveals about the human species. Whatever happened around 74,000 years ago, our ancestors came through it. The genetic evidence, whatever its precise cause, suggests the human population at some point was very small. Small enough that a modest amount of additional bad luck could have ended the lineage entirely. And it didn’t.