The new numbers, based on 2002 data, are higher than previously reported.
According to Lee Grossman, President of the Autism Society of America, "This report validates what we've been saying for quite some time, in that the numbers of those with autism has been under-reported. It also validates... that autism doesn't discriminate."
In addition to finding more children with autism than anyone thought, the study sheds light on another huge problem: Children are being diagnosed too late, usually not until they start school, reports CBS News medical correspondent Jon LaPook.
"This data today shows we're going to need more early intervention services and more therapists, and we're going to need federal and state legislators to stand up for these families," said Alison Singer, spokeswoman for Autism Speaks, the nation's largest organization advocating more services for autistic children.
To get a clearer picture of autism in America, the Centers for Disease Control and Prevention conducted a comprehensive review of the medical and school records of children in 14 states, reports LaPook.
The CDC study calculated an average autism rate 6.6 per 1,000. That compares with last year's estimated rate of 5.5 in 1,000.
However, the states studied are not demographically representative of the nation as a whole, so officials cautioned against using the results as a national average. The study doesn't include some of the most populous states like California, Texas and Florida.
Also, the study does not answer whether autism is increasing a controversial topic, driven in part by the contention by some parents and advocates that autism is linked to a vaccine preservative. The best scientific studies have not borne out that claim.
"We can't make conclusions about trends yet," because the study's database is too new, said Catherine Rice, a CDC behavioral scientist who was the study's lead author.
Autism is a pervasive development disorder usually not diagnosed in children until after age 3. The disorder involves delays in the development of many basic skills, most notably the ability to socialize or form relationships with others as well as the ability to communicate and to use imagination, according to WebMD. The cause is not known.
Studies show that the earlier the intervention, the better the outcome.
"I think that the most important thing is that parents [and physicians] are informed," says Dr. Alfredo Lowe, a Psychologist in New Jersey. "The more knowledge we have, the more power we have. The sooner that these kinds of disorders can be diagnosed, the better the care."
Deena Chackes was diagnosed early, at only 14 months, LaPook reports. Her family moved from Georgia to enroll her in The McCarton School in New York City. She's now doing "fantastic." "She talks, not always easy to understand, but she talks. She behaves like any other kid," her family told CBS News.
Scientists have been revising how common they think the disorder is. Past estimates from smaller studies have ranged from 1 out of every 10,000 children to nearly 1 in 100.
Last year's estimate of 5.5 out of every 1,000 U.S. children was based on national surveys of tens of thousands of families with school-age kids. That fit into a prevalence range found in other recent studies.
The CDC has been developing an alternate way of measuring autism prevalence, building a network of university and state health departments for ongoing surveillance of autism and developmental disabilities. The study released Thursday is one of the first scientific papers to come out of that effort.
"This is a more accurate rate because of the methods they used," said Dr. Eric Hollander, an autism expert at New York's Mount Sinai School of Medicine.
The study involved 2002 data from parts or all of 14 states: Alabama, Arizona, Arkansas, Colorado, Georgia, Maryland, Missouri, New Jersey, North Carolina, Pennsylvania, South Carolina, Utah, West Virginia and Wisconsin.
Researchers looked specifically at children who were 8 years old that year. They said most children with autism are identified for medical or educational services by that age.
The researchers checked health records in each area and school records when they were made available, looking for children who met diagnostic criteria for autism. They used those numbers to calculate a prevalence rate for each study area.
The rates varied from 3.3 per 1,000 in the study site in Alabama, which was made up of the state's 32 northernmost counties, to 10.6 in the site in New Jersey, which involved four counties, including metropolitan Newark.
Researchers say they don't know why the rate was so high in New Jersey. They think the Alabama rate was low at least partly because researchers had limited access to special education records there.