It was all laid out in a Dec. 7, 2006, memorandum from Peacock, the deputy administrator, to Gray, the EPA science adviser.

Now, the EPA would give the panel a "policy assessment" reflecting the agency's views. The policy assessment would be short on science and long on policy options, giving the administrator a wide berth in setting standards.

"It was done to give the administrator a way to control the process," said panel member Bruce Lanphear, a toxicologist and lead expert. Lanphear said he was so disgusted by what he saw as the undermining of the panel's role that he later moved to Canada.

The policy assessment would be published in the Federal Register, and the EPA science panel would have to stand in line and comment on it just like the general public.

Thus, reduced to the role of an outside critic, Henderson's panel would no longer shape and strengthen the EPA staff's scientific findings.

In justifying the move, Peacock said the process would help improve efficiency, while ensuring that the best science from a broad field of experts would be considered.

The panel was outraged. They were, after all, a creation of Congress, chartered under the Clean Air Act to serve as an objective scientific body in advising the EPA administrator.

They demanded the paper.

Then they got a stroke of luck.

The lead review had been prompted by a federal lawsuit, which meant the EPA would need permission from the presiding judge to alter the process.

The coalition objected to the changes, and were joined by the State of Missouri.

"Certainly, in this case, after a draft Staff Paper has already been completed, it seems as if the agency is trying to hide the ball, not bring it into the light," Missouri Attorney General Jeremiah W. Nixon wrote in a brief he filed in the case.

"Missouri believes EPA's new procedure is deeply flawed and actually undermines the interests it purports to serve."

The court ruled that the EPA would have to turn over the staff paper, as well as the policy assessment.

That allowed a unique opportunity to compare the two and measure what effect the new process would have on future reviews.

The results were alarming.

"If it had not been so serious, it would have been laughable," Henderson said. All 23 scientists on the lead review panel expressed concern over the regressive nature of the assessment.

The panel wrote Johnson in January to say the agency's revised process "represented a remarkable weakening of the scientific foundation of the . . . review process."

For example, the policy assessment said Johnson should view as acceptable blood-lead levels in children considered so toxic by the Centers for Disease Control and Prevention that they require treatment. The panel in a letter to Johnson in January called such a starting point "particularly troubling."

Jonathan Samet, chair of the epidemiology department at Johns Hopkins University, wrote that the policy assessment was "an unacceptable result of a process that should be evidence-based but has clearly gone off course and lost its links to a substantial body of evidence."