When did people reach the Americas?

In 2021 a study reported human footprints preserved in lake-margin sediments in New Mexico, dated by seeds in the layers above and below them to between about 23,000 and 21,000 years ago.

That is during the last glacial maximum, when the ice sheets were at their largest and the corridor through them was closed. If the dates hold, people were south of the ice while the standard route was shut, and had come before the ice closed it.

The dating was challenged on the grounds that the seeds were aquatic plants, which can take up old carbon from the water. The team returned with pollen from terrestrial plants and with a separate method dating the quartz grains themselves, and reported that both bear the original dates out. The quartz dates are minimum ages, and put the layer at about 21,000 years or more.

The argument has not stopped. A 2024 study of the same plant growing today argued the footprints could be at least 7,500 years younger, and a 2025 study of the ancient lake’s muds, led by one of the original authors, supported the glacial age.

Genomes of living and ancient Native Americans point to a founding population related to northeast Asians, isolated for some thousands of years before expanding south.

Modelling of ancient genomes puts the split between the northern and southern branches of Native Americans at 17,500 to 14,600 years ago, and probably south of the ice sheets. That sits awkwardly with footprints several thousand years older. If the footprints are as old as their dates, either their makers left no genetic descendants, or the genetic dating is missing something.

Monte Verde, in southern Chile, was excavated from the 1970s and reported dates around 14,500 years ago, a thousand years before the accepted arrival and far to the south, in northern Patagonia.

In 1997 a panel of specialists visited the site and accepted its upper layer. Later work dated nine species of seaweed from that layer, carried from beaches and estuaries and used for food and medicine, and reported scattered stone tools, bone and burnt patches in older layers, between about 18,500 and 14,500 years ago, which the excavators call probable rather than certain.

That acceptance, widespread by the early 2000s, is what ended the old framework. The footprints in New Mexico arrived into a field that had already learned to take an early date seriously.

In 2026 the site was put in question again. The first independent investigation of it in nearly fifty years argued, from dates on the surrounding stream deposits and a volcanic ash of 11,000 years ago lying beneath the archaeological layer, that it cannot be older than the middle Holocene, 8,200 to 4,200 years ago. Another team replied that the argument is flawed and that decades of work support about 14,400 years.

The corridor through the ice began to open about 15,000 years ago and was fully open by about 13,800, but the plants and animals that would have fed anyone walking it arrive in the sediment record only around 12,600. The coastal alternative, moving south along the Pacific edge, is favoured by many archaeologists and hard to test: global sea level at the glacial maximum was 134 metres below today’s, and those shorelines now lie tens to hundreds of kilometres offshore.

Much of the evidence that would settle this is under water, on drowned landscapes that five archaeologists writing in 2017 called “one of the last frontiers for archaeology in the Americas”.

The field spent decades rejecting claims of early sites, and many of those rejections were sound: doubtful dates, objects that may not have been made by people, layers whose context was never clear.

That history makes the bar for a new claim high, which is appropriate, and it also means a real result has to survive an unusually hostile review. The footprints are currently in the middle of that process, and since 2026 so is Monte Verde.