Three years ago, on a mid-Pacific island landing strip cut from a single swath in the coconut trees, an Australian filmmaker named Dennis O'Rourke climbed from a small airplane and began looking around to see what he might film. The island was Rongelap, in the then-U.S.-administered Marshall Islands, and the Australian government-sponsored Film Australia had commissioned O'Rourke to prepare six educational documentaries on Pacific island life.

O'Rourke knew the islands well; he had spent 10 years filming Pacific culture, and there were lessons he had learned about easing his cameras and his large Caucasian presence into communities largely unaccustomed to either. One moved more slowly, kept from stepping physically too close to people, defused the sexual threat by pulling out snapshots of one's wife and children. He was not certain just what he was after, but when the airplane broke down and his two planned Rongelap days stretched to 10, O'Rourke found himself listening again and again to the men and women who stopped their fishing and coconut gathering long enough to tell him a story that seemed to have the makings of an extraordinary film.

The tale O'Rourke was hearing, the essence of which is now compressed into his startling and award-winning documentary "Half-Life," was this: Thirty-two years ago, as part of its unprecedented series of Cold War-era nuclear tests, the United States dropped a hydrogen bomb -- the largest ever detonated on the surface of the Earth -- on the island of Bikini, 100 miles from Rongelap. The Bikinians had been evacuated, but the people in Rongelap had not. That evening their children, who had seen pictures of snow, played in the white powder that fell to the ground around them. "When I bailed out my canoe," an aging man tells the camera in slow Marshallese, "the water had turned yellow. And that night we ate the fish, and they tasted bitter."

"The stories that people were telling me about what had happened before, during and after the tests -- I knew I had to stay and make a film," O'Rourke says, leaning across the kitchen table in the house where he is staying as the San Francisco Film Festival screens "Half-Life." He is blond, broad across the shoulders and slightly rumpled-looking in his tweed jacket and day's growth of beard; O'Rourke is spending nearly every day and evening now discussing or arranging distribution for his documentary. "I just knew what the essence of the film would be," he says. "It would be a film which dealt with people living in a postnuclear environment."

O'Rourke knew a little about the Marshalls before he arrived, he says; he was familiar with the well-publicized plight of the Bikinians, who had lost their homeland to the American tests, and he knew there were angry accusations about American responsibility for the bombing's aftereffects on other island people. So many lawsuits have been filed, in fact, that seven American law firms now work jointly on the Marshall Islands Atomic Testing Litigation Project, which represents 13,000 people suing U.S. agencies for wrongful death, personal injury and destructive contamination of their property during the 12 years of U.S. nuclear testing in the Pacific.

But what O'Rourke says he found, as he set to work on his film, was something he insists seemed beyond belief in the early days of his research. The central and intensely controversial argument of "Half-Life" is that American officials, in their quest for data on the effects of nuclear contamination, deliberately allowed the islanders of Rongelap and nearby Utirik to be exposed to radiation so intense that as a people, O'Rourke says, they have now been "genetically destroyed" -- left with leukemia, miscarriages, birth defects, thyroid tumors and damaged genes that O'Rourke says will now be passed, "like Darwinism in reverse," from each generation to the next.

"I was unwilling to believe what the thesis of the film has become," O'Rourke says. "I'm not a conspiracy theorist-type person, and I'm not an antinuke person. I've never marched. I'm not an activist . . . I didn't come at this film from any polemical point of view. In fact, having seen the whole canon of antinuclear films, I find most of them to be quite ineffective in the end, because they present such a closed point of view."

"Half-Life," which in January received the Director's Award for Extraordinary Achievement at the United States Film Festival in Park City, Utah, uses archival footage to present what has consistently been the official American explanation of the aftereffects of "Bravo," the March 1954 bomb detonation that remains the largest of all America's nuclear tests. An American official, speaking calmly into movie cameras of the 1950s, announces that the predicted wind patterns should have carried bomb fallout to the uninhabited area north of Bikini -- but that the wind shifted south, leaving Rongelap and Utirik at the edge of the wind-borne contamination.

Using government documents and filmed interviews with three American weather and radio technicians, O'Rourke suggests in "Half-Life" that the official account is wrong -- that by midnight the night before the tests, military personnel knew the direction of the wind and proceeded anyway. Island inhabitants had been evacuated for earlier, less powerful nuclear tests, the film declares; this time they were neither removed nor warned about possible dangers of contact with the contamination.

And although government records showed military ships and helicopters were waiting in the immediate vicinity, O'Rourke says, it was two days before the people were taken from the fallout-contaminated land. "The dose that is killing the Marshallese was a cumulative dose -- it took 50 hours," O'Rourke says. "Any of those ships, especially the one that was right there -- they knew exactly where the people were living. Instead of removing them right then -- and they could have removed them at minimal risk to the servicemen . . . the people at that stage would have all been rescued, and the dose they would have received would have been very much the same dose you and I might receive if we had been working in a nuclear plant for a year."

Thus American scientific and military officials provided themselves with an ideal test population to examine the effects of nuclear fallout, O'Rourke says -- particularly when they sent the Marshallese back to what O'Rourke says were still-contaminated islands, and then scheduled regular medical and scientific examinations that have continued to this day. "These people are a perfect control group," O'Rourke says. "They're not going anywhere. They're a closed gene group. They intermarry. Their diet is closed -- they can't just go down to the Safeway and get food from elsewhere. And there's nowhere else for them to come into contact with carcinogens."

"Half-Life" premiered only three months ago in the United States -- the film was introduced last summer in Melbourne and has since been shown in a number of international film festivals -- and because distribution will not bring it to Washington and other American cities until this fall, few scientists or government officials have seen O'Rourke's documentary so far. A spokeswoman for the Defense Nuclear Agency, which provided O'Rourke with some of his documents and archival footage, says DNA personnel have seen the film and publicly wish to comment only briefly. "While we found the film to be interesting and informative, we strongly disagree with the implication that the responsible government officials intended to engage in human experimentation," she says.

"It's absurd . . . absolutely ridiculous," says Herbert York, director of the Institute on Global Conflict and Cooperation in San Diego, a physicist who in 1954 directed the Livermore laboratory and visited the Marshalls shortly after the Bravo test. York has not seen "Half-Life," and was not among the scientists O'Rourke interviewed, but he says he cannot accept either the decades-old memories of American weathermen or the modern-day conclusions about what precautions should have been taken when.

"What we knew about the wind on those days made us think it was safe," York says. "In my experience the people did the best they could . . . They thought it was going to go slightly northeast, and the past experience was just not all that good."

As for the delay in removing the islanders, York says, "There was general stupefaction and surprise . . . it really took time for them to figure out what had happened. And even if they did figure out what happened, it took time for them to figure out what to do about it. And I don't think there's anything more complicated or sinister about it. The notion that it was deliberate is just totally bizarre . . . An individual might react quickly, but a bureaucracy never does."

O'Rourke says he has heard these explanations before, and indeed that they helped persuade him at first that the Marshallese must be mistaken when they told him they believed they had been used, as his filmed American radio operator puts it, as "guinea pigs." "There's a lot of guilt floating around out there in people who were responsible for some of these things in the early years," O'Rourke says. In his film, he says, "there is no one document -- nor do I believe there's one document that exists -- that says, 'Right, these people will become an object of study' . . . Circumstantially, I think the case is won. But the words that I use to this day are, 'Decisions were made deliberately to allow them to be exposed . . . Decisions were made which in the end resulted in great damage to these people -- the death of these people.' "

O'Rourke calls his film, in the opening credits, "A Parable for the Nuclear Age," and it is as parable, he says, that he refrains in the documentary from presenting much present-day political information about the Marshalls. The film does not discuss the Compact of Free Association, which was signed by President Reagan last January and must be approved this summer by councils within the United Nations; the document essentially severs the United States' 39-year-old trusteeship of the Marshalls and other Micronesian islands that have now become independent republics. Included in the compact is a clause that grants $150 million to settle all future Marshallese claims arising from the testing. That clause is now under legal attack by attorneys for the Marshallese, since it would exempt the United States from any responsibility beyond the $150 million, which one attorney estimated might be a fifth of the total claims now pending.

And the medical stories of nuclear aftereffect, in their jarring juxtaposition against blue ocean waters and large-eyed island children, are presented not as documented statistics but as personal and sometimes bewildered accounts. An old man turns the pages of a photo album, pointing to the snapshots of his teen-age son being peered at by American doctors and then lying in a flower-covered coffin. A gray-haired woman, with no audible emotion in her voice, describes the pregnancy that began when she returned home after the "Bravo" test: "My stomach started to swell, but before the proper time I gave birth to something which I cannot describe. It did not look human. I don't know how to say it . . . like the innards of a beast. It could not survive -- it was dead at birth."

O'Rourke chose to present his dramatic material this way, he says, partly because he believes mortality and illness statistics to be suspect -- "any figure I quote is going to be somebody's figure," he says. More powerful, to him, were the voices of the angry islanders, like the woman near the end of his film who describes her son's slow death before American doctors who poked at him, she says contemptuously, as though he were a chicken being prepared for dinner. "Don't Americans know that every life is precious?" the woman says. "I used to think they're smart, but now I think they're crazy . . . They're smart at doing stupid things."

And O'Rourke does not pretend to be a scientist or an investigative reporter, he says -- he is a filmmaker, who prefers to make what he calls "observational films, about these people and the pressures being applied to them." Some of his previous films have examined the independence of Papua New Guinea, or the arrival of television on the island of Yap; his wife, in fact, is a New Guinea woman who met O'Rourke while he was making one of his documentaries.

His wife stayed at the family home in Canberra while he was filming in the Marshalls, O'Rourke says, and worried about him. "When we first went there, we were planning to have another child, and my wife was terrified because I was there," he says. He did submit to a full medical examination, and was told that he seemed to have been eating a lot of contaminated coconuts. But their child was born entirely healthy, O'Rourke says, and he is convinced the tradeoff was worth it. "I just considered that comparatively speaking, any risk I took was minor," he says.