Military servicemen who witnessed atmospheric tests of atomic bombs in the 1950s don't have an increased risk of death, or of death from cancer, compared to people who were in the armed services at the time but weren't near the explosions.

That's the main conclusion of a study, seven years in the making, released by the National Academy of Sciences' Institute of Medicine (IOM) yesterday.

The "atomic veterans," as they're sometimes called, do have a somewhat larger-than-expected rate of leukemia, a blood-cell cancer that can be caused by radiation. However, the increase is small--about 25 "excess" cases of leukemia in a group of about 68,000 servicemen studied--and may have occurred simply by chance.

The question of whether thousands of soldiers were put in harm's way during the Cold War has provoked bitter debate since 1976, when a veteran dying of leukemia in Idaho asserted his illness stemmed from radiation exposure suffered during a 1957 bomb test in Nevada.

The question has also been unusually hard to answer. The National Academy of Sciences oversaw a study in the 1980s that was officially withdrawn in 1992 after researchers learned the Defense Department's roster of exposed soldiers was highly inaccurate. The new study, done by the IOM's staff at the cost of about $4.3 million, and with a much more exhaustive search for test witnesses, replaces it.

"This study is the biggest, and really the best," said Harvey Checkoway, an epidemiologist at the University of Washington who helped advise the IOM researchers.

The findings, however, are unlikely to change some minds.

"I'm not convinced at all," said Robert Campbell, a 67-year-old veteran from Portland, Me., who watched several explosions from Eniwetok atoll, where he was an Army clerk for 13 months. He's done his own enumeration, and believes there are more cancer deaths than the epidemiologists found.

The study's conclusions will have no immediate effect on policy. The Department of Veterans Affairs already has a list of diseases it presumes to be "exposure-related" for purposes of compensation, if a veteran can prove he was exposed to a specified minimum dose of radiation.

About 200,000 Defense Department personnel were exposed to atmospheric radiation between 1945 and the end of above-ground testing in 1963. In most cases, they were soldiers and sailors on military outposts or ships who were miles from the detonation sites.

The new study looked at the experience of 68,168 people who attended five series of tests--three in the South Pacific, and two in Nevada--where 62 bombs were exploded between 1951 and 1957. Their experience was compared to that of a group of 65,000 people, matched for age, military occupation, pay-grade and other variables.

About one-quarter of the people in each group are dead. The researchers located the death certificates of about 92 percent of them, tabulated the causes, and compared the findings.

There was no increase in mortality in the bomb-test witnesses. That group, however, had a slightly higher rate of death from accidents and suicide, and a slightly higher rate of death from leukemia, although the latter was not statistically significant. The increased rate of leukemia death was found only in the Nevada witnesses. Why that was so--or whether it happened simply by chance--is unknown.

The study also found a slight increase in nasal cancer--a very rare disease--and prostate cancer--a very common one--in the bomb witnesses. Neither cancer is thought to be related to radiation exposure.

"The set of leukemia findings is consistent with the results of other studies of military participants in nuclear tests, and consistent with the hypothesis that these are radiation effects," wrote Susan Thaul, the IOM epidemiologist who directed the study. "The other findings listed are more likely to be chance occurrences."

The researchers had hoped to correlate mortality and radiation exposure, with the latter obtained from monitoring badges some soldiers wore at the time, or from estimates made later based on units' proximity to the blasts and weather conditions at the time. However, they concluded the data that exist were not good enough to support such a "dose-response" analysis, so one wasn't done.

Studies of the survivors of the Hiroshima and Nagasaki bombings that ended World War II found increased rates of leukemia and, to a lesser degree, cancer of other organs, such as thyroid, lung, breast, esophagus and colon.

A study of the health of children who were exposed to the bomb blasts as fetuses found no increase in cancer deaths or other deaths, through 1985. In general, scientists have been surprised at how little the long-term health of Japanese bomb survivors seems to have been changed by their exposure.

CAPTION: The study looked at the experience of 68,168 people who attended five series of atomic bomb tests--three in the South Pacific, and two in Nevada--between 1951 and 1957.