The results alarmed her.

Dr. Petreas and a colleague examined tissue samples from harbor seals and women's breasts for traces of polybrominated diphenyl ethers, or PBDEs.

Between 1989 and 1998, levels of the chemical in the seal samples had increased 100-fold. PBDE levels in the breast tissue were 10 to 40 times as high as those found in an earlier study in Sweden.

"They were so high we thought we made an error," says Dr. Petreas, a
53-year-old specialist in toxic pollutants with the California Environmental Protection Agency.

Dr. Petreas's research is on the forefront of a rapidly escalating debate
over PBDEs. Their fire-retardant properties are believed to have saved thousands of lives, but PBDEs have been shown to be harmful to animals in laboratory tests. There's still no proof that they cause problems in humans.

What is clear is that levels in humans are rapidly building up -- apparently doubling every two to five years.

That gap between their potential harm and the lack of conclusive proof has created rifts between state and federal officials and even within the ranks of the U.S. Environmental Protection Agency. Not wanting to wait for federal action, California in August became the first state to ban two PBDE compounds that have been shown to accumulate in humans.

The ban will take effect in 2008.

PBDEs have also become the flashpoint of a larger debate in Europe and the U.S. over the tens of thousands of chemicals that came into use before environmental laws were tightened in the 1970s. Though U.S. companies have agreed to test 3,000 high-use chemicals by 2005, other chemicals that predate the Toxic Substances Control Act of 1976 have been largely ignored.

Even under the 1976 act, industry testing is voluntary, with U.S. companies required by law only to disclose to the EPA any available data on the harmful effects of their chemicals.

Prodded in part by public concern over recent environmental issues such as Mad Cow disease in Britain and the dioxin that turned up in Belgian chicken feed, the European Union has proposed a controversial measure that calls for industry to evaluate some 30,000 previously untested chemicals for potential health and environmental hazards by 2012. Of particular concern are chemicals that persist in the environment, such as PCBs, or polychlorinated biphenyls, and PBDEs.

Fearing that the EU legislation could become a global standard, the Bush administration is pushing hard to limit its scope. Charles Auer, director of the EPA's Office of Pollution Prevention and Toxics, says widespread testing isn't necessary when the bulk of chemicals in use are relatively safe. He says a system focusing on a smaller number of chemicals that are more likely to be hazardous could save time and money -- in addition to sparing laboratory animals. "There are simpler, cheaper ways to meet the same needs," says Mr. Auer.

In the absence of mandatory testing and regulation, what chemicals get
studied and how debate over potentially hazardous chemicals plays out is often left to chance and the interests of researchers. The story of PBDEs shows this circuitous path.

Developed in the early 1970s, PBDEs are used as fire retardants in products from foam cushions to the plastic casings of computers and televisions. The chemicals were first discovered in the environment downstream from an industrial plant in Sweden in the early 1980s. Since then, they have been found just about everywhere scientists have looked for them: in sewage sludge in the U.S., in cormorants in England and in beluga whales in the Canadian Arctic.

Scientists theorize that the chemicals get into the environment when they are being incorporated into products as well as when those products are being used or disposed of. In the case of foam in furniture, for example, scientists believe that the chemicals leach into the air as the foam breaks down over time. They are then carried by the wind and deposited around the world by rain and snow, working their way up the food chain and accumulating in predatory animals and people.

PBDEs are currently manufactured by Great Lakes Chemical Corp. of
Indianapolis; Albemarle Corp. of Richmond, Va., and Dead Sea Bromine Group, a unit of Tel Aviv-based Israel Chemicals Ltd. About a quarter of the 239,000 metric tons of the PBDEs produced globally in 2001 were used in the U.S. The global PBDE market is valued at about $774 million.

Three Compounds

There are three main PBDE compounds. The California ban affects just two, which are used in furniture and some plastics. The third, which is used in plastic casings of computers and televisions, hasn't been shown conclusively to accumulate in people. The third type, however, represents more than 95% of the PBDEs used each year.

Peter O'Toole, U.S. program director for the Bromine Science and
Environmental Forum, which represents the three PBDE manufacturers, says the industry acknowledges that levels of the two compounds banned in California are rising in people, but it maintains they aren't toxic. He says the industry is spending several million dollars in testing PBDEs, especially to demonstrate the safety of the compound used in most plastic casings.

Evidence that PBDEs are potentially harmful to people dates back to the mid-1980s, when scientists at the National Toxicology Program of the National Institute of Environmental Health found that mice fed high doses of PBDEs were prone to liver cancer. But because the doses fed to the mice were so high compared with the levels found in humans at the time, the data were largely dismissed.

Then, a 1998 study of Swedish breast milk, which showed that PBDE levels were doubling every two to five years, reinvigorated the research effort. In 1999, researchers at Sweden's Uppsala University fed 10-day-old mice the PBDE mixture used in furniture and found the animals' learning, memory and behavior were impaired. Similar findings were made by the EPA and at the University of Rome.

The question facing scientists is at what point do PBDEs affect health?
Newborn mice show developmental disorders at concentrations of PBDEs that are 10 times as high as those currently found in the most highly exposed parts of the U.S. population. Some scientists say that isn't a comfortable margin, especially considering the rapid rate of accumulation in humans. "These are levels approaching the developmental disorders in mice," says Linda Birnbaum, the EPA's director of experimental toxicology.

Mr. O'Toole of the bromine-industry group agrees that the high levels of
PBDE compounds are a concern. But he says scientists are wrong to
extrapolate from the results of studies on laboratory animals. "That's not viable," he says

Dr. Petreas was in Sweden at a conference on hazardous chemicals when the breast-milk data were first disclosed. "Typically the research reported at these meetings is mundane," says Dr. Petreas. "This was something spectacular."

Back home, Dr. Petreas and a colleague, Dr. Jianwen She, didn't have a ready source of breast-milk samples, so they settled on studying seal samples from San Francisco Bay and breast tissue available from a breast-cancer study.

Dr. Petreas's boss at the time, Kim Hooper, wanted to make sure their data didn't get ignored. A scientist with an activist's bent, Dr. Hooper, 63, proposed that his colleagues wage a campaign within the California EPA to raise awareness among officials about PBDEs. "Otherwise it's going to blow up in their face," he said.

But Dr. Petreas was reluctant to be seen as political. She kept a distance when Dr. Hooper attended a scientific "share fair" at the California EPA's Department of Toxic Substances Control headquarters in Sacramento in November 1998. He took every opportunity to lobby his superiors to note the Swedish breast-milk study, even scrawling on poster board: "Did you ever wonder if the chemical that was in your TV set would end up leaching from the landfill into the groundwater, end up in your food and finally in your breast milk?"

Later that summer, Dr. Petreas and Dr. She reported their findings to an annual dioxin conference in Monterey, Calif. A paper in the scientific
journal Chemosphere followed. Meanwhile, Dr. Hooper and a state EPA colleague, Thomas McDonald, published an article in Environmental Health Perspectives in May 2000 that synthesized some 50 reports on the potential hazards of PBDEs and called for more research and caution.

As the weeks wore on, Dr. Hooper began to chafe at what he felt was the slow-moving bureaucracy of the California EPA. "I'd send up bulletins and they would disappear into the void," he says. William Rukeyser, a spokesman for the California EPA, says such criticism is like "dialing 911 and wondering why the fire truck isn't there yet."

In the meantime, Dr. Petreas pursued another study, this one comparing blood samples from two groups of California women, one of them archived from the 1960s and another from the 1990s, as well as fatty tissue from another group of 1990s women. She found very high levels of PBDEs in all but the 1960s samples and concluded that the increasing levels of PBDEs, "particularly in young women of reproductive age, pose a potential public health threat to future generations."

Though PBDEs were becoming the buzz in scientific circles in the U.S. and abroad, there was still little movement in policy-making circles in
Sacramento.

Task Force

That began to change when Dr. Hooper and Dr. McDonald invited environmental advocates and industry representatives to join an interagency task force they had convened on PBDEs. Soon an advocacy group called Environment California had drafted a measure to ban PBDEs and signed on powerful lawmakers to champion it. "We laid out what we think the science says," says Dr. Hooper. "They took it and ran with it."

The campaign in California also got a boost when Europe banned two of the three main PBDE compounds earlier this year. The ban goes into effect in 2004.

The original California bill, which sought to ban all three of the major
types of PBDEs by 2006, generated serious opposition from electronics makers and others who thought the deadline was too tight. The final version bans just the two compounds used mostly in furniture, beginning in 2008.

Great Lakes Chemical, the only U.S. company to manufacture the two types of PBDEs under the ban, didn't fight the final legislation, saying the 2008 ban date gives the company adequate time to produce alternatives. Already, furniture companies such as Swedish manufacturer Ikea have stopped using PBDEs as flame retardants, as have computer makers such as Apple Computer Corp. and Hewlett-Packard Co.

Now, California EPA officials are growing restless with their colleagues in Washington. "Chemicals of this sort should be regulated at the national level," says Winston Hickox, California EPA secretary. Absent a national approach, "the state must protect its citizens' health," he says.

Some EPA scientists agree. What is known about other chemicals that build up in the body, says Dr. Birnbaum, the EPA toxicology expert, "is that we don't want them around."

While Mr. Auer of the U.S. EPA acknowledges that rising levels of PBDEs present "red flags," he says that regulating the compounds will require more testing. "We're not at the point to be able to do that yet," he says.