← All articles

Mikhail Polyakov

How many people lived in the Americas before Columbus: why estimates ranged from 1 to 112 million and how charcoal from ancient hearths is narrowing them down

Level: BeginnerReading time: 8 min

Topics: History, Science

An excavation with dark patches of ancient hearths and bits of charcoal, glass vials with charcoal, a brush and a trowel nearby, hills and earthen mounds in the distance

Short answer: there is no exact number, and there may never be one. Estimates of the Indigenous population of North and South America in 1492, according to Live Science, ranged from 1 to 112 million people. A 2019 synthesis gave 60.5 million for both Americas, with an interquartile range in that same work of 44.8 to 78.2 million. For what is now the US and Canada, experts interviewed by Live Science give 2 to 4 million. The most unexpected part: according to radiocarbon dates, the population of North America was at its largest around 1150, several centuries before Columbus, and then declined across a large part of the continent. Below is how this was calculated and where these numbers are weak.

Why counting is so hard

There are no good exact counts: "the numbers are all over the place," anthropologist George Milner tells Live Science. On October 12, 1492, Columbus landed on the island now called San Salvador and met the Lucayan people there. In his diary he wrote about a "great multitude of people." Within a few decades after that, the Indigenous population of the Americas fell sharply. According to the calculations of the authors of the 2019 work, European epidemics killed about 90% of the Indigenous population over the following century.

So scientists estimate indirectly: from how many people the land could feed, from historical records, from traces of settlements. Each method has its own error, hence the spread of more than a hundredfold.

112 million and why it was abandoned

One of the first modern researchers to try to count the population was anthropologist Henry Dobyns. In a 1983 book, he relied on two things: how many people each type of landscape could feed, and historical records of 16th-century epidemics. The result was about 112 million people, more than a hundred times higher than some earlier estimates.

Milner, an anthropologist at Penn State, explains to Live Science what was weak about this count. Dobyns, in his words, filled the landscape with people up to the limit it could feed, so that all environmentally similar places ended up equally populated. And he assumed that epidemics spread widely and universally, which only makes sense if the landscape is fully settled. Excavations in the 1970s and 1980s showed that the Americas were settled unevenly. Therefore, according to Milner, Dobyns' estimates turned out to be much too high.

Regional estimates

Next, scientists began estimating the population of individual regions from traces found in excavations: from the density of pottery sherds at a site, the number of stone tools and flakes, the number and size of houses, and the ages at which people whose skeletons were found had died.

In 2019, the journal Quaternary Science Reviews published a work by Alexander Koch, Chris Brierley, Mark Maslin and Simon Lewis that brought such regional estimates together. Live Science writes about 129 estimates from 82 published studies; the abstract of the paper itself refers to 119 published regional estimates. The authors divided the Americas into seven regions: the Caribbean, Mexico, Central America, Inca territory, Amazonia, North America and the rest of the continents. The result: 60.5 million people in 1492. The abstract gives an interquartile range for this number of 44.8 to 78.2 million: the middle half of the values obtained in the calculation falls within these limits, without the lowest and highest quarters. For North America, meaning present-day Canada and the US, the work gave 3 to 6 million, that is, no more than a third of Dobyns' estimate of 18 million. The most populous peoples of the Americas in 1492, according to this work, were the Aztecs, mostly in what is now Mexico, and the Inca on the west coast of South America.

Earlier, in 2010, Milner and geographer George Chaplin took eastern North America, where archaeology is best studied. They compared historically known group sizes with the sizes of places where people lived and estimated where and how many people lived in clusters. For the eastern US and southeastern Canada, this came to about 1.6 million people. For all of North America in the early 16th century, before European diseases sharply reduced the population, they gave 2.8 to 5.7 million, with 3.8 million as their best estimate.

A new method: counting by charcoal from hearths

Archaeologist Robert Kelly of the University of Wyoming tells Live Science that comparing regions requires a measure that works everywhere. Such a measure was found in fires: everyone built them, for heat and for cooking, and charcoal from hearths can be radiocarbon dated. The method shows the time range in which the firewood was most likely burned.

The idea of the count is simple. The more people lived in a given period, the more charcoal, bones and plant remains they left behind to be dated later. If you add all such dates up over time, you get a curve: where there are many dates, there were most likely more people.

Dates from North American excavations are collected in the Canadian Archaeological Radiocarbon Database (CARD). Geographer Matthew Peros of Bishop's University in Quebec and co-authors, as Live Science reports, combined the date ranges from this database into a single curve in 2009. At that time the database had about 25,000 records. According to the curve, the population of North America grew quickly from about 50 BC and reached its peak around AD 1150. From then until contact with Europeans, it may have been declining. One possible explanation Live Science gives: people began living more densely, and more of them died from diseases such as tuberculosis. Peros admits that the peak around 1150 surprised him too: he had expected to see it around 1500.

In 2025, Kelly, Madeline Mackie, Spencer Pelton and Erick Robinson published a work in the journal PNAS based on the database, which by then, according to Live Science, had grown to more than 100,000 dates. What it shows, according to the abstract:

  • the Indigenous population of North America over roughly the last 14,000 years was largest around 1150, then declined, briefly recovered shortly before 1500, and after that the number of dates drops sharply;
  • the authors tested whether the peak around 1150 could be explained by how samples were chosen for dating, and rejected that explanation;
  • across 18 river watersheds, peaks fall at different times, from about 800 to after Europeans arrived, and in most watersheds in the interior of the continent the decline began roughly between 1080 and 1300;
  • in a large part of the territory, around the Great Lakes, in New England, on the Mid-Atlantic coast, on the Central Plains, in the Northwest and in California, the population did not decline until Europeans arrived.

From the curve to the number of people

The curve shows when there were more or fewer people, but not how many. To get a number, as Peros explains, you need to set a starting population and an average annual growth rate. The Peros study used this assumption: about 15,000 years ago a few large families, around 1,000 people, entered North America. With it, the result was about 2.5 million in 1150 and about 1.8 million in 1492.

An important caveat from the same Live Science article: if people arrived significantly earlier than 15,000 years ago or in greater numbers, the figures would be higher. For example, traces at the White Sands site in New Mexico are dated to 23,000 years ago. So 1.8 million is the result of a specific assumption, not a measurement.

Kelly says that had Europeans arrived around the year 1000, they would have met a much larger Indigenous population, and history could have turned out very differently. For comparison, Live Science cites the city of Cahokia near present-day St. Louis: at that time it covered about 16 square kilometers, and nearly 20,000 people lived in it.

Why this matters

Milner says that a full estimate for all the Americas may be reached someday, but he himself will most likely not see it: the work is long, laborious and not flashy. He believes that the demography of the past helps us understand how population and nature changed over long periods, and that this will be useful for decisions about the future. Peros adds something else: North America was not an empty continent when the Europeans arrived.

How to read news about the size of ancient populations

This is my advice, not conclusions from the article.

  1. Look for a range, not a single number. The "60.5 million" in the abstract comes with a range of 44.8 to 78.2 million. If a news story gives only one number, it is worth finding the original source.
  2. Look at the method. An estimate from how many people the land could feed, an estimate from traces of settlements and an estimate from radiocarbon dates rely on different things and can differ greatly.
  3. Separate measurement from assumption. The radiocarbon curve shows rises and falls, and a number of people is obtained from it only with assumptions about the starting population and growth. Change the assumption, and the number changes.
  4. Check which territory is meant. 60.5 million is both Americas, 2 to 4 million is the US and Canada, 1.6 million is only the eastern US and southeastern Canada. They are easy to mix up in retellings.

Summary

The number of Indigenous inhabitants of the Americas in 1492 is still known only as a range. The 1983 estimate of 112 million turned out to be too high, the 2019 synthesis gave 60.5 million for both Americas, and for the US and Canada the experts interviewed by Live Science give 2 to 4 million. Radiocarbon dates showed that the population of North America was largest around 1150, and in most interior regions the decline began long before Europeans arrived. As Milner said, long and laborious work lies ahead.

I work on AI agents and automation, and here I sometimes write about science. If you want to look at my projects or discuss your own task, visit my portfolio.

Sources