ASHINGTON, Jan. 10 - In an eagerly awaited report on perchlorate, one of the most controversial unregulated toxic pollutants in the country's drinking water and food supplies, the National Academy of Sciences said Monday that people would be safe if exposed to daily doses 20 times those under consideration by the Environmental Protection Agency.

Depending on how federal and state regulators interpret the academy's recommendation, the Defense Department, its contractors and other federal agencies responsible for contamination from perchlorate, a component of solid rocket fuel, could avoid cleanup costs of hundreds of millions of dollars.

The environmental agency and the states of California and Massachusetts have already taken the initial steps in the regulatory process, with the E.P.A. and Massachusetts both suggesting a maximum safe level of one part per billion, and California setting a goal of six parts per billion. Thus far, no regulation on the maximum safe level of perchlorate in drinking water has been made final.

Large doses of the chemical, in widespread use by the Defense Department since the 1950's, have been shown to inhibit the thyroid gland's ability to take up iodide from a person's diet. Insufficient iodide has been linked with impaired neurological development, but the report said that the evidence the panel examined "is inadequate to determine whether or not there is a causal association between perchlorate exposure and adverse neurodevelopmental outcomes in children."

The scientists on the National Academy panel avoided arriving at a figure for safe drinking water levels, saying that was not their charge, recommending instead a safe level based on body weight. Some state regulators and representatives of environmental groups, doing their own extrapolations from the panel's report, said it would support a drinking water standard of 20 parts per billion. Others said the conclusions could support maximum levels of less than three parts per billion.

Pentagon scientists, using the same human studies that underpinned the academy's report, had concluded that the maximum safe level of perchlorate in drinking water supplies was 200 parts per billion.

Groundwater around the country has been found to contain trace levels of perchlorate. The chemical has been detected in the Colorado River, a water source for 15 million people in the Southwest. The town of Bourne on Cape Cod closed some wells because of high perchlorate levels.

But perchlorate's toxicity is hotly disputed, as are safe exposure levels. The debate led four federal agencies, including the Defense Department, to ask the academy to assess perchlorate's adverse health effects.

In its report, the 15-member panel, led by Dr. Richard B. Johnston Jr. of the University of Colorado School of Medicine, said that risk assessments should be based on human studies that indicate when the thyroid's uptake of iodide is inhibited. The E.P.A.'s 2002 risk assessment had relied in part on studies of rats that indicated changes in some brain structures after perchlorate exposure.

Scientists from environmental groups like the Natural Resources Defense Council and the Environmental Working Group pointed out that the main human study involved only seven healthy adults who ingested scaled amounts of perchlorate for 14 days.

Regulators in both California and Massachusetts said Monday that they would review the report and, if necessary, adjust their preliminary findings on perchlorate. Allan Hirsch, a spokesman for the Office of Health and Hazard Assessment in California, said that changes might not be necessary, adding that the National Academy dose recommendation "is highly consistent with the calculations we made."

In an earlier call with reporters, officials of the Natural Resources group said that the evidence considered by the panel had been unfairly weighted on the side of industry and the Defense Department, which, along with the White House had an undue influence on the process. They cited evidence they obtained under the Freedom of Information Act, which showed extensive e-mail communication among high-ranking administration members about the charge given to the academy panel.

Richard Canaday, a representative of the White House Office of Science and Technology Policy, responded in a telephone interview: "There is no basis for that claim. This is an attempt to distort the science by attacking the process." The academy, Mr. Canaday added, is the "gold standard of independent scientific review."