This suggests that current Environmental Protection Agency guidelines on mercury exposure should not apply to the type of mercury in vaccines -- and could help answer doubts about the safety of some vaccines, the researchers wrote in this week's issue of Environmental Health Perspectives.

"The current debate linking the use of thimerosal in vaccines to autism and other developmental disorders has led many families to question whether the potential risks associated with early childhood immunizations may outweigh the benefits," Thomas Burbacher of the University of Washington in Seattle and colleagues wrote in their report.

The Institute of Medicine, an independent body that advises the federal government, has said there is no evidence of any link between vaccines and autism. It has advised researchers looking for the causes of autism to look elsewhere.

This has enraged autism activists, who fear a cover-up.

Burbacher's team said it would make sense to study more closely the effects of thimerosal, a mercury-based preservative once widely used in vaccines. It has been removed from most childhood vaccines because of the furor.

Thimerosal remains in the influenza vaccine and in some vaccines used outside the United States.

Current government advice on vaccine exposure limits are based on studies done on people who were exposed to methyl mercury from industrial accidents. It can clearly cause long-term brain and nervous system damage.

An earlier study calculated that children receive 187.5 micrograms of ethyl mercury from thimerosal-containing vaccines given over the first 14 weeks of life. This can exceed EPA guidelines for methyl mercury exposure during pregnancy.

Some experts have argued that thimerosal breaks down into a different form of mercury in the body -- ethyl mercury -- and that this is cleared more quickly.

Tests on human infants suggest this does happen, but their brains cannot be directly examined.

Burbacher's team tested 41 newborn monkeys, feeding them either methyl mercury or giving them shots of thimerosal in doses comparable to those given vaccinated human infants. The mercury from the vaccines was cleared out of the body much more quickly than was methyl mercury, they said.

It took just over eight days to completely clear mercury from thimerosal, while it took 21 days to clear methyl mercury from the blood, they found.

"Brain concentrations of total mercury were significantly lower by about three-fold for the thimerosal-exposed infants when compared to the methyl mercury infants," they wrote.

The researchers said this does not mean thimerosal is harmless and urged more research.

"This information is critical if we are to respond to public concerns regarding the safety of childhood immunizations," they wrote.