“This is a very significant study because it confirms that genetic factors are involved in the cause of the disorder,” said Dr. Peter Szatmari, a leading autism researcher who is the head of child psychiatry and behavioral neuroscience at McMaster University in Ontario. “But it shifts the focus to the possibility that environmental factors could also be really important.”

As recently as a few decades ago, psychiatrists thought autism was caused by a lack of maternal warmth. And while that notion has been discarded in favor of genetic explanations, there has been growing acceptance that genes do not tell the whole story, in part because autism rates appear to have increased far faster than our genes can evolve.

“I think we now understand that both genetic and environmental factors have to be taken seriously,” said Dr. Joachim Hallmayer, an associate professor of psychiatry and behavioral sciences at Stanford and the lead author of the new study, which is to be published in the November issue of Archives of General Psychiatry.

Other experts have cited factors like parental age, multiple pregnancies, low birth weight and exposure to medications or maternal infection during pregnancy.

In the new study, the largest of its kind among twins, researchers looked at 192 pairs of identical and fraternal twins whose cases were drawn from California databases. At least one twin in each pair had the classic form of autism, which is marked by extreme social withdrawal, communication problems and repetitive behaviors. In many cases, the other twin also had classic autism or a milder “autism spectrum” disorder like Asperger’s syndrome.

Identical twins share 100 percent of their genes; fraternal twins share 50 percent of their genes. So comparing autism rates in both types of twins can enable researchers to measure the importance of genes versus shared environment.

The study found that autism or autism spectrum disorders occurred in both children in 77 percent of the male identical twins and in 50 percent of the female identical twins. As expected, the rates among fraternal twins were lower: 31 percent of males and 36 percent of females.

But surprisingly, mathematical modeling suggested that only 38 percent of the cases could be attributed to genetic factors, compared with the 90 percent suggested by previous studies.

And more surprising still, shared environmental factors appeared to be at work in 58 percent of the cases.

“We, like everyone else, were very surprised because we didn’t expect it to be that high,” said a senior author of the study, Neil Risch, a geneticist and epidemiologist at the University of California, San Francisco.

The rate of autism occurring in two siblings who are not twins is much lower, suggesting that the conditions the twins shared in the womb, rather than what they were exposed to after birth, contributed to the development of autism.

A second article, also released early on the journal’s Web site, found an elevated risk of autism in children whose mothers took a popular type of antidepressant during the year before delivery. But the authors reassured women taking these drugs — so-called S.S.R.I.’s like Prozac, Zoloft, Celexa and Lexapro — that the risk was still quite low: 2.1 percent in children whose mothers used them in the year before delivery, and 2.3 percent in the first trimester of pregnancy.

Dr. Joseph Coyle, the editor in chief of the psychiatry journal, called the two studies “game changers.”

Clara Lajonchere, an author of the twin study and vice president of clinical programs for the research and advocacy organization Autism Speaks, said that “much more emphasis is going to be put on looking at prenatal and perinatal factors with respect to autism susceptibility.”

She added, “We need to not just study the environmental factors, but the relation between the genes and the environment.”

“For pregnant women or those thinking about having a family,” she said, “prenatal care is critical, and if a pregnant woman is taking any kinds of medication, she should work closely with a physician.”