Since the mid-1990s, there has been a struggle within New World Paleo-Indian studies about the population of Native Americans in the Americas prior to Columbus. I finally tackled an entire book on the subject recently ... "1491" by Charles Mann. 

In a nutshell, there are low-counters and high counters within the anthropological sciences. For decades, the weight was clearly on the side of low-counters - those who thought only a few million  (single digits, usually one-to-two million) Native Americans were in North America before Columbus arrived. Several discoveries in the last decade changed that... and now the weight has shifted to the high counters - those believing there were as many as 25 million (making it nearly as populous as Europe by some accounts).

There are several reasons new discoveries have shifted opinion.

1. Diseases that we expected to take a toll on Native American populations after Columbus probably had an even greater effect than anthropologists first thought. There is convincing evidence that:

* The entire Cape Cod coastline was decimated prior to the landing of the Pilgrims. This we know from written accounts from Pilgrims based on Squanto - the Indian who had been captured several years before and then returned to Cape Cod the year before the Pilgrims arrived.

* We also know Native American populations in the Caribbean were reduced dramatically after Columbus based on his own study.

* We know smallpox dramatically reduced The Inca population in South America in waves after Columbus. There was a huge outbreak prior to Pizarro's conquest.

* We know Cortez wasn't able to defeat the Aztec alliance until smallpox decimated them just before his final conquest. Smallpox then moved quickly from Mexico into the southwest U.S. and into the Rocky Mountains effecting the North American western tribes throughout the 1600s.

In the case of diseases, it never stopped prior to European arrival. Many of us have heard of the smallpox outbreaks during the Revolutionary War, but we almost never hear of other outbreaks like the influenza outbreak in the early 1700's that killed several important Indian chiefs nor do we hear about measles, cholera, typhus, and even hepatitis A outbreaks among Indians. This population decimation continued unabated even after the arrival of Europeans.

We now know through DNA that the Native American gene pool was even more susceptible to these diseases than first thought. It wasn't just a matter of prior exposure. We also know that Amerindians did not understand quarantine like Europeans did at the time. Family members would gather around their sick family members.

We know that in the 1500s, Hernando de Soto saw scores of Indian villages in his trek west through the American south. In one small strip along the Arkansas and Red Rivers he accounted for more than 50 Indian villages. After a gap of nearly 100 years, the area wasn't visited again by a European until LaSalle visited. He saw, in the exact same area, a total of 3 villages and dozens of empty villages. Interestingly De Soto had brought farm pigs on his early journey - a huge carrier of several bovine pox diseases.

There are no real hard figures for the population of North American Indians prior to Columbus that would be considered reliable. Baseline data of death rates changes the algebra dramatically. However, it is generally assumed now that the death rate was badly underestimated in the 20th century as it pertained to disease.

2. We also know that the idea of a single migration across the Bering "land bridge" about 12,000 years ago is probably not nearly as significant as we first thought. There's now good reason to believe people were here prior to 12,000 years ago. The best recent theories suggest waves of migrations - the land bridge scenario would be only one of many. No one is quite sure how the the migrations took place. Genetically, we are at least sure that Native Americans are related to Asian Siberians... so we know there were west-to-east migrations, but how that happened is a bigger mystery now than it was 20 years ago. We can throw out the textbook examples we grew up with.

Evidence uncovered recently in Monte Verde, Chile - of a Indian population pre-dating all known sites - was revolutionary. The site is still being studied, but most Paleo-anthropologists believe that the dates of about 11,000-14,000 years ago make the single migration theory almost obsolete. 

http://anthropology.net/2008/05/08/earliest-known-archaeological-evidence-of-americans-found-in-monte-verde-chile/

Also, archaeology in Peru, Bolivia, and Chile are now uncovering sites between 4000-6000 years old that show fairly dramatic technological advances and rather stable systems of government control. If you went back in time 5000 years, the biggest empire in the world at that time would not have been Egypt. It likely would have been western South America.

These population centers suggest a fairly widespread and stable population prior to the Spanish arrival. Even in the Amazon Basin, they are now uncovering artifacts and even bones

The biggest problem - one that has held scientific opinion in check for so many years - is the paucity of actual skeletal remains in the America's as opposed to the numbers found in Europe (where there are significantly more).

Scientists and anthropologists have still not gotten the proper number of skeletons to alter their understanding entirely. That's why there is still some scientific debate. However, other evidence is being found at such a high rate, that most professionals continue to go with higher numbers of Native Americans than first thought, more advanced civilizations than first thought, and dates that go farther back for migrations than first thought.

Finally. our understanding of Indians within the boundaries of the U.S. and Canada has been poorly interpreted.

In 1100 A.D., one of the largest communities of people in the temperate zone of either North America or Europe was Cahokia - an Indian village near St, Louis - a mound building community of at least 15,000 people and probably around 40,000. That would have been larger than either Boston or Philadelphia in the year 1700 - at least 600 years later.

http://en.wikipedia.org/wiki/Cahokia

We are also now much more aware of the importance of maize and how agriculture was a dominant part of society prior to the arrival of Columbus. It's much clearer now that harvesting crops, cleared fields, slash and burn control, and irrigation techniques were in common practice prior to Columbus. Descriptions by Europeans of open forests, almost like old English parks, and fields of maize harvested by eastern Indians suggest that the words "hunter-gatherer" societies probably don't accurately describe conditions about the land prior to Columbus. Indians were manipulating the environment in ways European historians have underestimated for centuries.

We do know that evidence for these practices were only partially observed by Europeans - probably because the population decimation between 1492 and 1620 (arrival at Plymouth) was so enormous that much of the agricultural ground went fallow. Amerindian societies were broken up and impoverished, Indian civil wars occurred, and new alliances for survival were made just prior to the arrival of the English.