Thousands of chemicals should be re-assessed for possible toxicity to human and environmental health, according to a new study.
Scientists found that conventional tests underestimate how some substances accumulate along the food chain.
These include some pesticides, and possibly some pharmaceuticals.
Writing in the journal Science, the scientists say about one-third of organic substances in commercial use would need re-testing.
That could amount to about 10,000 substances. The researchers expect most would turn out to be benign.
Last month, European Union legislation on the Registration, Evaluation, Authorisation and Restriction of Chemicals (Reach) came into force.
It will see about 30,000 chemicals in industrial use tested for health and safety impacts at a total cost of about 10bn euros ($13bn).
Something in the air
Substances are said to be "bio-accumulative" or "persistent" if they remain in the body after ingestion.
Concentrations rise higher up the food chain. A substance found at a certain concentration in plankton will be at a higher concentration in small fish that eat the plankton, still higher in big fish that eat the small fish, and higher still in bears or seals that eat the big fish.
Twelve types of toxic persistent organic pollutants (Pops) including DDT, dioxins and polychlorinated biphenyls (PCBs) have been banned worldwide under the Stockholm Convention (Pops Treaty).
Regulators routinely assess the bio-accumulative potential of a chemical using a ratio called the octanol-water partition co-efficient (Kow).
It is basically a measure of how effectively the substance dissolves in fat versus water. A partiality for water means the substance is likely to be transported out of a fish's body as it respires, while a fondness for fat means it is likely to remain.
This works well as a measure of accumulative potential in food chains involving fish, shellfish and plankton, the researchers say.
But they conclude that a different measure is needed for accumulation in air-breathing animals - a measure called Koa that looks at whether the substance in question can be lost across the lung membrane during respiration.
Off the radar
The researchers compared the accumulation in three different food chains of two chemicals: PCB-153 and beta-hexachlorocyclohexane (beta-HCH), a close relative of the well-known insecticide lindane.
One food chain goes from plankton to fish, the second from lichen to caribou to wolves, and the third from plankton through fish to walruses, seals and polar bears.
PCB-153, which has a high Kow, accumulated along all three food chains. But beta-HCH, with a low Kow but high Koa, only showed accumulation along the lichen/caribou/wolf and marine mammal food chains.
Its potential to accumulate would have been missed by conventional tests.
Classes of compounds possessing low Kow but high Koa include endosulfans and HCHs, which are used as insecticides, musk xylene, an ingredient of perfumes and soaps, and the tetrachlorobenzenes.
"About one-third of these organic chemicals are not on the radar screen," said Frank Gobas from Simon Fraser University in Burnaby, British Columbia.
"They won't all be bio-accumulative, but they all have the potential to be bio-accumulative."
Many may be effectively metabolised in the body and disposed of that way.
But Professor Gobas' team believes they should all now be examined using this new measure of bio-accumulation to see if there is a hitherto unexpected threat to health and environmental well-being.