According to most archeologists and geneticists, the best theory for how the first humans migrated to the Americas is the same one that many likely learned in grade school: they crossed the Bering Strait from Asia via a now-extinct land bridge.

“One can only imagine what they thought when they got to North America south of the ice sheets and looked around and realized nobody else was home,” says David J. Meltzer, an archeologist and professor of prehistory at Southern Methodist University. “‘Wow, we’ve got an entire land to ourselves.”

Thanks to advances in genome sequencing and data analysis, we know that some of the first humans to set foot in North America (known as Paleo-Americans) were direct descendants of ancient people in Siberia, which is solid evidence for the land bridge hypothesis.

But not everyone is convinced that all Paleo-Americans walked to the Americas from Asia. There are outlier archeological sites in both North and South America that date to times before a land route was accessible. Could some of the first Americans have crossed oceans to get here? And do those theories hold up to scientific scrutiny?

Here is the evidence for three theories explaining how the first humans arrived in America: the land bridge theory, the trans-Pacific migration theory and the controversial Solutrean hypothesis.

The theory with near-unanimous support from both archeologists and geneticists is that the first humans to populate the Americas arrived on foot via a temporary land bridge—across a region known as Beringia—that connected Eastern Siberia to Alaska for a span of roughly 5,000 years.

Why did they cross that forbidding, near-lifeless landscape into an unknown world? And once they arrived, why and how did they push southward so quickly? The classic explanation—that they were following animal herds—is far too simplistic, says Meltzer, author of First Peoples in a New World: Populating Ice Age America. “It’s not as though animals were headed down to Miami.”

Archeologists and anthropologists are still searching for the “why” of “how” of the first human migration to the Americas, but they’re zeroing in on the “when.” Thanks to a handful of well-preserved Ice Age archeological sites, and a precious few fully-sequenced genomes from Paleolithic human remains, a rough timeline has emerged.

The ancient ancestors of the first Americans left Siberia between 24,000 and 21,000 years ago. That’s been confirmed by comparing the DNA of Paleo-Americans with the DNA of Paleo-Siberians to pinpoint the moment when those two human populations diverged.

“We’re analyzing a mind-bogglingly huge scale of data to detect these patterns,” says Jennifer Raff, a geneticist and anthropologist at the University of Kansas. “These are really subtle differences that you wouldn’t pick up on if you were not sequencing every single base pair of these ancient genomes.”

According to paleoclimatologists, thick ice sheets covered much of the northern latitudes from 23,000 to 19,000 years ago, a period known as the Last Glacial Maximum. With all of that sea water trapped in ice, sea levels dropped, exposing a stretch of dry land between Asia and North America.

If a land route to the Americas was open at least 20,000 years ago (if not earlier), when exactly did the first Americans cross it? That’s one of the liveliest debates among archeologists.

For decades, says Meltzer, the consensus was that the first Americans arrived around 13,000 years ago. That date corresponds with thousands of finds across North America pointing to a shared archeological culture known as Clovis (named after a specific type of fluted spear point first found in Clovis, New Mexico in 1933). Clovis points were everywhere by 12,000 years ago, so the archeological evidence overwhelmingly backed an initial arrival 13,000 years ago.

But then came Monte Verde in Chile. At this remarkable site, thousands of miles south of the Bering Land Bridge crossing, archeologists unearthed solid evidence of a human settlement dating to 14,800 years ago, well before Clovis. After Monte Verde broke the 13,000 barrier, even more pre-Clovis sites emerged: Paisley Caves in Oregon (14,400 years ago), the Debra L. Friedkin Site in Texas (15,500 years ago) and Cooper’s Ferry in Idaho (16,000 years ago).

Those early, pre-Clovis dates were exciting, but they posed a new challenge for archeologists. According to geological evidence, thick glaciers blocked the land route from the Bering Land Bridge into North America until the glaciers started retreating around 15,000 to 14,000 years ago. So how were humans already in Idaho and Texas before the land route opened up? Or as far south as Southern Chile by 14,800 years ago?

“This is a point of great controversy among archeologists,” says Raff. “You’ll find plenty of archeologists who say [the first Americans] came through an ice-free corridor [over land], which didn’t open up until 14,000 years ago. But we have pre-Clovis sites that are early enough in the Americas that can’t be explained by the ice-free corridor.”

For Raff and others, a more likely scenario is that the pre-Clovis humans reached America in boats. The western coastline of North America was ice-free thousands of years earlier than the interior. The pre-Clovis humans could have crossed the Bering Land Bridge on foot, then used simple boats to skirt around the glaciers and make their way down the West Coast of North and South America.

Genomic evidence also supports the coastal boat theory.

“All of the ancient genomes that have been sequenced carry the signature of rapid movement—the rapid divergence of different lineages in the Americas—suggesting that people were moving very quickly,” says Raff. “It’s harder to move that quickly on foot than by boat.”

It could be, says Meltzer, that there were many “pulses” of migrations into the Americas via the land bridge over several millennia, but that the pre-Clovis arrivals were highly mobile and relatively few in number.

“We know that people can be present long before they pop up on archeological radar,” says Meltzer. “It may be that what Clovis represented [13,000 years ago] was the population in the Americas finally getting to the point where there were enough of them out there producing enough stuff that they became archeologically visible—really visible.”

In 2015, a geneticist named Pontus Skogland published some intriguing results from sequencing the genomes of modern-day indigenous people living in the Amazon region of South America. While the vast majority of their genome was shared with indigenous people in other parts of the Americas—and clearly descended from Eastern Siberia—there was a mysterious “signal” in the data.

According to that signal, the indigenous Amazon villagers appeared to share an ancient common ancestor—named “Population Y”—with indigenous people from “Australasia,” a region that includes indigenous Australians, New Guineans, Papuans and more.

For some, this was evidence that ancient inhabitants of Australia or its nearby islands may have used Polynesian-style wayfinding to sail thousands of miles across the open Pacific Ocean to arrive on the coast of South America.

While a trans-Pacific migration to the Americas sounds plausible, there are problems with the theory.

“We don’t have any evidence, archaeologically or genetically, of a trans-Pacific migration,” says Raff. “We do have evidence of a faint signal of shared ancestry between some South Americans, both ancient and modern, and individuals in Australasia, but it doesn’t match what you would expect from a trans-Pacific migration. The signal would be a lot stronger in individuals on the West Coast, and less as you move farther East, but it doesn’t fit with that model. It’s scattershot throughout the population.”

Because of the nature of the signal, Raff and other geneticists believe that Population Y is very, very old, and originated in Asia (the same genetic signal was present in a 40,000-year-old man found in a Chinese cave). Tens of thousands of years ago, some descendants of Population Y went north and others went south. Some of those northern descendants ultimately crossed the land bridge and made it to South America, while some of the southern descendants populated Australasia.

There are no archeological sites in the Americas that pre-date 16,000 years ago, according to Meltzer. “There are claims to sites that are 20,000, 25,000 or even 130,000 years old,” says Meltzer, “but at the moment those claims are highly questionable at best.”

Not everyone agrees. Bruce Bradley, an emeritus archeology professor at the University of Exeter, thinks that there’s compelling evidence that humans were occupying sites along the East Coast of North America as far back as 20,000 years ago. And he has a controversial theory for how they got there so early: they came from Europe.

Bradley’s theory—which he detailed in a 2013 book, Across the Atlantic Ice: The Origin of America’s Clovis Culture (co-authored with the late Dennis Stanford)—centers on something called the Solutrean archeological culture. Roughly 23,000 to 18,000 years ago, ancient humans living in modern-day France and Spain produced a distinctive and elaborate toolkit of stone blades, spear throwers and harpoons that archeologists call Solutrean.

“Everybody agrees, Solutrean-age humans were incredibly innovative,” says Bradley. “They lived in a relatively harsh environment [during the Last Glacial Maximum], but also started looking toward the ocean. We’re starting to see the beginnings in Western Europe of maritime adaptations—looking at the ocean and other aquatic resources.”

Bradley thinks that Solutrean hunters and fishermen may have fashioned simple “skin boats” (similar to the Inuit umiak) to expand their hunting territories along the ice-clogged North Atlantic. Eventually, those travels took them further and further across the ice until they arrived on the eastern shores of North America.

“We’re not talking about Solutreans getting in boats and sailing across the ocean,” says Bradley. “It wasn’t a migration, but an expansion of hunting territory.”

For Bradley, the evidence that Solutreans landed in America is found in sites like Parsons Island, Maryland, where stone blades and other tools (tentatively dated to more than 20,000 years ago) bear a striking resemblance to Solutrean technology.

“If you put this group of artifacts and technologies at a site in France, you wouldn't even question it,” says Bradley. “It’s Solutrean, period.”

Bradley hypothesizes that these ancient Solutreans were some of the earliest (if not the first) humans to arrive in the Americas and that their technology was what became Clovis, the archeological culture that spread across North America by 13,000 years ago.

The Solutrean hypothesis has many critics, Meltzer and Raff among them. Meltzer wonders how an ancient people with no archeological evidence of boat-making could have navigated an ocean. “Look, the Titanic didn't make it,” he says. “How are a bunch of Solutreans in a boat going to cross the Ice Age North Atlantic?”

For Raff, the proof (or lack of it) is in the genetic record. In 2014, scientists sequenced the genome of the Anzick child, the remains of a Clovis-era boy in Montana who lived 12,700 years ago, making him the oldest burial in the Americas.

“The Anzick genome showed absolutely no genetic evidence of European ancestry, nor do any genomes of pre-contact Native Americans,” says Raff. “Anzick very roundly refuted the Solutrean hypothesis.”

For his part, Bradley isn’t conceding. He says that the Solutrean hypothesis was never meant to “replace” the Bering Land Bridge theory.

“People definitely came out of Siberia; there’s no question about that,” says Bradley. “We’re not saying that everybody came from what’s now southwestern Europe. The Solutreans were just one of probably multiple groups that came to the New World at various times and from various places.”