The new study found children and teens exposed to high levels of bisphenol A, or BPA, were more likely to be obese.
"Clearly bad diet and lack of exercise are the leading contributors to childhood obesity, but this study suggests a significant role for environmental, particularly chemical factors in that epidemic," said study author Dr. Leonardo Trasande, an associate professor of pediatrics and environmental medicine at the NYU School of Medicine.
Trasande and colleagues measured body mass and urinary BPA -- an indirect way of measuring BPA exposure -- in more than 2,800 American children and teens. While over 92 percent of the study subjects had detectable levels of BPA in their urine, those with the highest levels were 2.6 times more likely to be obese than those with the lowest levels, even after controlling for diet and exercise.
The findings, published today in the Journal of the American Medical Association, add weight to calls for a broader ban on BPA in food packaging.
"I applaud [the Food and Drug Administration's] decision to ban BPA in baby bottles, but this ban did not go far enough," said Oladele Ogunseitan, chair of population health and disease prevention at the University of California, Irvine, who was not involved with the study. "There are viable BPA chemical alternatives that should be researched more. It is time to ban BPA exposure in the general population, including adults."
BPA, an ingredient in some hard plastics and metal cans, made headlines in 2008 when it was found to leach out of plastic when heated. Laboratory studies in cells and animals have linked the chemical to cancer, infertility, diabetes and obesity. But the health consequences of chronic, low-level exposure in humans remain unclear.
"The FDA ban of BPA in baby bottles is not based on definitive scientific studies," said Dr. Robert Brent, professor of pediatrics, radiology and pathology at Thomas Jefferson University in Philadelphia. "The country is bordering on lunacy from the exaggerated fear of chemicals."
The FDA admits it "sees substantial uncertainties with respect to the overall interpretation of many published studies, and, particularly, their potential implications for human health effects of BPA exposure," according to a statement. But the agency said it will consider Trasande's study in its "ongoing evaluation of the safety of BPA."
The link between BPA and obesity is biologically plausible, according to Trasande. In animals, the chemical makes fat cells bigger and inhibits the function of adiponectin, a protein that helps break down sugars and fats. It also appears to disrupt hormones that play a key role in energy balance.
"Our study can't identify obesity as being caused by BPA. But in the context of increasing evidence from experimental studies, it raises further concern," he said.
More than one-third of American children are overweight or obese, according to the U.S. Centers for Disease Control and Prevention, and childhood obesity has more than tripled over the past three decades.
But the American Chemical Council, a plastics trade group, said attempts to link the nation's obesity problem to BPA distract from "real efforts" to address obesity.
"Due to inherent, fundamental limitations in this study, it is incapable of establishing any meaningful connection between BPA and obesity," spokesman Steven Hentges said in a statement. "In particular, the study measures BPA exposure only after obesity has developed, which provides no information on what caused obesity to develop."
Indeed, it could be that obese kids are simply more likely to consume canned food or products packaged in BPA-containing plastic, according to the study.
"Association does not prove causation," said Charles Santerre, professor of food toxicology at Purdue University in West Lafayette, Ind. "America is obese, not because of BPA, but because we consume more calories than we burn."
ABC News' Dr. Chandani Patel contributed reporting.