Yet a review of the data offers little support for Johnson's claim.

Roughly speaking, the data form a straight line on a graph: the cleaner the air during the years studied, the fewer people died.

And statistically speaking, that evidence is just as reliable at 13 or 14 as it is at the higher, average level that guided Johnson - as measured by the width of what are called confidence intervals.

C. Arden Pope, who co-authored both the big studies the panel relied on, was puzzled by Johnson's focus on the average.

He said it was a bit like looking at a group of people, a few of whom smoke two packs a day while most don't smoke at all. Under Johnson's logic, Pope said, it would be like concluding that it is safe to smoke the "average" amount, say, 10 cigarettes a day.

"It's silly," said Pope, a professor of economics at Brigham Young University.

Yet Pope said he sympathized with the administrator, because improving air quality to conform with the Clean Air Act's "adequate margin of safety" is likely to be unrealistic.

"You've got the empirical evidence that looks like almost any pollution will hurt you," Pope said.

How clean is clean enough? While lawyers may debate the meaning of "adequate," Johnson's own staff assembled some numbers: how many lives would be saved in each case.

Such estimates are tricky, because no one knows how low you can go and still achieve additional health impacts. Experts consulted by the agency said Johnson's rule would prevent anywhere from 1,200 to 13,000 premature deaths each year, whereas tightening the annual standard by just one microgram would prevent somewhere between 2,200 and 24,000.

Jason Burnett, a former EPA official involved in the rule-making process, said Johnson's decision to leave the standard alone was in keeping with the signals coming from the White House.

"They didn't want to tighten it, even though the vast majority of the evidence and the scientific advice we were receiving was to tighten it," said Burnett, who left the agency in part because he disagreed with the fine-particle rule.

When pressed, he declined to reveal details of when and how the pressure was applied.

But several months before Johnson issued his decision, the White House OMB wrote him to express skepticism about the big studies upon which the science panel based its advice. In an April 2006 letter, the OMB's Donald R. Arbuckle, a senior official, urged further analysis and "a more complete characterization of the uncertainty."

The new rule on particles has now been challenged in federal court by the American Lung Association, with a decision expected early next year. If the rule stands, the incoming Obama administration still could elect to revisit and toughen the standards, but the process could take years.

Through a spokeswoman, the EPA declined to comment on Johnson's reasoning for leaving the annual pollution limit unchanged, citing the litigation. Under Bush, the agency did move to reduce particle pollution from buses and construction equipment. But legally speaking, the annual pollution limit is the driving force for cleanup efforts by the individual states, and by leaving it alone, Johnson left little impetus for change.

For Henderson, what Johnson did was worse than a missed opportunity. Johnson's rule, she felt, did not provide the margin of safety required by the Clean Air Act.

"His decision contrasts with what the law clearly says," she said.

Like a sunburn

In her lab in Albuquerque, N.M., Henderson was disappointed after the 2006 decision.