Dioxin -- the chemical that has come to symbolize the dangers of industrial pollution -- appears to be less dangerous than previously thought, according to one of the most comprehensive epidemiological studies of the contaminant ever conducted.

Using the health records of 5,000 chemical industry workers exposed to the chemical over the past 40 years, government scientists were able to find a "slight" excess cancer risk only among workers with dioxin levels absorbed by their tissues some 500 times higher than normal.

The results suggest that public concern over the levels of dioxin typically found in the environment may be largely unfounded. It also appears to bolster the growing view of many scientists that U.S. policy toward the chemical, a common byproduct of industrial processes notorious for its presence in the defoliant Agent Orange, are far too strict and that millions of dollars are being wasted in its unnecessary regulation.

"This is very reassuring," said George Carlo, chairman of the Health and Environmental Sciences Group and one of the nation's leading dioxin experts. "Here we have the group of people who have been exposed to higher levels of dioxin than perhaps anyone in the world, followed for a longer period of time than any other study, and it is not bearing out the cancer risk hypothesis. If we were going to see something we would see it here in spades. And we haven't."

The question of dioxin's safety has been a subject of debate among scientists and environmentalists since tests of the chemical on rats and mice in the 1970s showed it to be, for them, a potent carcinogen.

Since that time, however, two developments have led many scientists to revise their earlier opinion of dioxin. The first is an analysis and reinterpretation of the original animal tests, which showed that the conclusions drawn from the data may have been overstated. The second is that repeated studies of people exposed to unusually high levels of dioxin have not confirmed the picture of the chemical painted by the animal tests.

For example, even after years of monitoring, no appreciable dioxin-linked health effects have been found in Vietnam veterans who came in contact with dioxin contained in the defoliant Agent Orange used in that war.

The study of chemical industry workers, conducted by scientists at the National Institute for Occupational Safety and Health and published in today's issue of the New England Journal of Medicine, appears to confirm this view.

After examining medical records of the 5,000 chemical factory workers, the scientists concluded that only those exposed to massive amounts of dioxin suffered any ill effects, and those effects formed only a modest indictment against the chemical.

For example, the average exposure of workers to dioxin was 2.7 years, but the average time they spent working in chemical plants was 12.6 years. This raises the possibility, the researchers said, that dioxin may have been only one of many potentially harmful substances that the workers had come into contact with during their careers. The "study could not completely exclude the possible contribution of other occupational carcinogens or smoking . . . ," they said.

This possibility, some experts said, was also suggested by the puzzling pattern in cancers observed among the workers. No increases in liver cancer were found in the group, despite the fact that the liver is the organ known from animal tests to be the primary target of dioxin. What the study did find was an excess in lung cancer, which had not previously been suspected to be a tumor caused by dioxin. The researchers said that it is conceivable that the lung tumors were caused by the exposure of some workers to asbestos or smoking.

Despite the study's qualifications, however, many dioxin experts welcomed it as providing the first definitive evidence that risk from the chemical is linked to the amount of exposure.

The study found increased cancer risk only among the about 1,500 workers who had what are estimated to be the highest exposure levels in history -- 500 times the norm. No increased risk, by contrast, was found among the other 3,500 workers even though they had, on average, dioxin levels 90 times higher than normal.