One worker collapsed on the factory floor, his body ravaged by lymphoma. Two others died within 105 days of different forms of leukemia. By the time Challie Freeman came down with a rare bone disease in the fall of 1979, questions had morphed into suspicions:

Was something at the U.S. government's uranium plant making workers sick?

One possible answer--radiation exposure--seemed persuasive to Freeman's doctor. He fired off a letter to the Paducah Gaseous Diffusion Plant. "It is imperative," he wrote, "that we learn as soon as possible the extent, nature and type of radiation to which he was exposed."

The reply--"no significant internal exposure"--was brief and emphatic. It was also false.

While the plant was denying knowledge of significant hazards to Freeman's doctors, confidential records showed the opposite: Freeman had tested positive multiple times for exposure to radioactive uranium and had even been restricted from working around uranium, an internal company memo shows.

In August, The Washington Post reported that Paducah workers were unwittingly exposed to highly radioactive plutonium and neptunium on the job from the 1950s to the 1970s. A subsequent four-month Post investigation has found additional evidence that plant officials kept employees uninformed about chemical and radiation hazards. In some cases, such as Freeman's, the plant withheld accurate medical information on radiation exposure--even while it privately tracked cancer deaths among workers.

A limited review of Paducah employee death records also turned up rates of leukemia among workers that appear higher than normal, based on government mortality statistics. Epidemiologists who reviewed the findings described the data as intriguing but cautioned that a much more intensive scientific study was needed, involving investigators with full access to employee records and medical histories, to establish whether a pattern existed. Such a study has not been done at Paducah.

The 48-year-old uranium plant is the subject of an Energy Department investigation into worker health and safety practices. Union Carbide Corp., which allowed its operating contract to expire in 1984, declines to comment, saying its Paducah managers are long gone from the company. Energy Secretary Bill Richardson, whose agency owns the facility, has apologized for the failure to disclose plant hazards and has promised compensation for sick workers.

Any outside attempt to review medical issues at Paducah is complicated by a lack of complete information. The Energy Department, citing privacy laws, declined to release lists of workers and their assignments. But The Post obtained company rosters listing more than 200 Paducah employees who were hired to work in some of the plant's most dangerous uranium-handling areas between 1951 and 1971. Scores of death certificates were examined and more than 120 surviving employees who worked in those areas were interviewed.

Professional help was retained to categorize deaths, and a software program developed by the National Institute for Occupational Safety and Health was used to compare incidences of cancer to national rates.

The result: The incidence of leukemia at Paducah appeared elevated, according to epidemiologists who reviewed the data. Of the 211 people on the lists who could be located--about 13 percent of the plant's work force in an average year--10 died of cancers of the blood and lymphatic system, including six of leukemia. By comparison, government mortality statistics suggest that only a single leukemia death would be expected in a group of adults of that size.

Cancer clusters are difficult to document, and cancers are not necessarily caused by radiation. Some studies at other Energy Department plants have suggested links between workplace hazards and cancers; others have not. Whether chronic exposure to low doses of radiation causes cancer has been hotly debated for decades.

Still, several epidemiologists who reviewed the results said the unusual incidence of leukemia and other rare diseases suggests the need for a closer look.

"The findings are interesting and noteworthy and are grounds for a more complete study of the question," said David Richardson, an epidemiologist who is researching radiation health effects for the World Health Organization.

Senior Energy Department officials said the findings highlight a major policy dilemma for the agency: whether to pursue more studies or to expand pilot programs to directly compensate workers who get sick. Yesterday, the department announced that it had shifted spending priorities in its fiscal 2000 budget to increase money for health studies and medical monitoring at Paducah. However, officials worry that studies may not be the right approach.

"Epidemiology is not going to answer the questions precisely enough," said David Michaels, an epidemiologist and the assistant energy secretary for environment, safety and health.

Energy Secretary Richardson said he has proposed legislation to change the way his agency deals with its sick workers.

"Instead of fighting claims, we're actually helping workers without the debate about the rates of illness," he said. "The legislation we sent to Congress takes the burden of proof off those who are sick."

Documents obtained in October under the Freedom of Information Act show that Union Carbide began tracking the repeated cancer cases in its work force in the 1970s.

The first to die was Wade McNabb, a 20-year veteran who succumbed to chronic leukemia in 1972. That same fall, another worker died of multiple myeloma, a bone marrow disease.

Alton Henson died of leukemia in 1976. Two years later, three workers--Arvil Bean, Leonard Lindblad and David Wilson--died of leukemia or bone marrow diseases within a span of six months.

By 1982, the company had counted 13 fatal cancers of the blood or lymphatic system out of a relatively stable work force that ranged from 1,200 to 2,000 people. The list appears on a single sheet of paper--stamped "confidential" and copied to senior plant officials--identifying workers sometimes by initials. How Union Carbide intended to use the list is unclear, but the plant's records show no attempt by contractors to investigate possible links between the deaths and workplace hazards.

Meanwhile, plant workers were told everything was fine. When Challie Freeman fell ill with his deadly bone marrow disease at 59, plant officials offered a lot of sympathy but little truth, family members say.

Responding to a hematologist's queries about possible radiation exposure, a plant physician in a letter described Freeman as a "very fine man" whose exposure to hazardous materials had been near zero. Medical records produced by the plant showed "no significant internal exposure," based on years of weekly urine tests for uranium.

Not until 15 years after his death in 1984 did family members obtain his medical records from the Energy Department and learn the full story: Company tests had indeed found high levels of uranium in his body in the 1950s--so much, in fact, that Freeman once had to be moved to a different work area. His widow, Sue, recalls that he was transferred to a different job in the 1950s after being told simply that his urine was "hot."

Freeman's physician, Nashville hematologist John Flexner, remembered that the company's response "downplayed the exposures."

"They made you think there was no way this could be a case of cause-and-effect," Flexner said. "I guess I was naive to think they were telling the truth."

Union Carbide said that it did not have the ability to respond in the Freeman case because of the 20-year passage of time.

Plant policies required that workers exposed to certain amounts of radiation be moved to other, less hazardous jobs. But new records show this was ignored in some cases in which workers received up to twice the maximum dosage.

One who never got the word was A.B. Burris, a 74-year-old retiree who learned of his past exposures when he asked the Energy Department for his medical files this fall.

"They say I was put on 'strict restriction,' but I never found out about it until weeks ago," he said. "I can tell you they never changed my job or said anything to me about it."

Workers knew even less about potentially deadly plutonium and neptunium that spread through the plant in shipments of recycled nuclear reactor uranium fuel from the 1950s to the 1970s, plant documents show.

Confidential, 40-year-old memos released by the Energy Department in September showed that Union Carbide officials had decided against testing workers for exposure to the radioactive metals because of fears that workers would "use it . . . as an excuse for hazardous-duty pay."

Newly released memos show that senior managers were aware of the plutonium and neptunium problem as early as 1959 but concluded in classified studies that contaminants were not a health hazard because the amounts in each shipment were small--a maximum of 10 parts per billion of plutonium in each uranium shipment.

But over the years, the two metals began accumulating in soil and waste materials.

In a survey of Paducah plant buildings conducted in the early 1990s, more than half of the work areas sampled exceeded the plant's safety limits for plutonium and neptunium--in some cases by a factor of 10. A survey of a men's locker room found high levels in shower stalls and even on toilet seats.

Workers did know enough about radiation hazards to formally request additional safeguards.

When Union Carbide decided to stop providing mechanics with coveralls, the plant's union demanded in 1986 that the company take responsibility for "radiation carried into our homes, autos and other areas." Union Carbide denied the request, although in 1975 the union negotiated the right to protective clothing on demand.

The union was less successful in efforts to secure workers' rights to take regular breaks in a radiation-free lunchroom. In a written grievance in 1979, the union said workers "should not have to eat in a contaminated area."

The company denied the request.

Ailing workers in the past have had difficulty proving harm because they lacked accurate monitoring data, David Fuller, president of the Paducah chapter of the Paper, Allied-Industrial, Chemical & Energy Workers Union, testified at a Senate hearing on Paducah in October.

While applauding government promises to financially aid ailing Paducah workers, Fuller and other union officials called for a compensation program for all workers that "reverses the burden of proof onto the government" while expanding medical monitoring for those most at risk.

"Monitoring is imperative," Fuller said, "but without any other remedy, monitoring is simply a process to watch people get sick and die."

Director of computer-assisted reporting Ira Chinoy, database editor Sarah Cohen, and staff researchers Alice Crites, Nathan Abse and Nancy Shiner contributed to this report.

Challie Freeman

Job: Cascade worker, security officer

Age at death: 64

Illness: Myelofibrosis

Did radioactive exposure on the job make Challie Freeman sick? His doctor suspected a link, but plant managers said no. Asked by doctors to provide details of Freeman's work history, a Union Carbide memo described light exposure to the skin but "no significant internal exposures."

Fifteen years after Freeman's death, the family obtained confidential plant memos that showed the opposite: Freeman had been restricted from uranium work in the 1950s because of "repeated positive urine samples" for radioactive uranium. The uranium remained high after weekends away from the job, the memo said.

Freeman became sick from a slowly progressing bone marrow disease in the 1970s and died in 1984. Near the end his weight plummeted from 190 pounds to 100 and he was in constant pain, said his wife, Sue, who quit her job to care for him. `We always wondered if it was the plant that made him sick,` she said. `Now I have no doubt.`

David R. Wilson

Job: Cascade operator

Age at death: 54

Illness: Lymphosarcoma

Like most Paducah workers, Wilson said little about his job, though sometimes he'd confide to his wife when he was exposed to unusually high levels of radiation. "He would say just he had been `hot,'" remembers his widow, Winnie. One day in early 1978 he was rushed to the hospital after becoming ill at work. Tests confirmed he suffered from a form of lymphoma, which ended his life just four months later.

Wade McNabb

Job: Cascade operator

Age at death: 55

Illness: Leukemia

The doctor's eyes spoke volumes. After breaking the awful news to McNabb -- a diagnosis of leukemia at age 40 -- he asked the ailing man where he worked. The reply, "Atomic Energy Plant, Paducah," prompted a nod and a knowing look. "Oh, yes," the hematologist said, "I'm treating several patients from Oak Ridge," Paducah's sister plant in Tennessee. McNabb began treatment and returned to the same job to preserve his salary and health benefits. "We didn't know what else to do," Dove, his widow, says. "You couldn't even talk about it at work, not if you wanted to keep your job."

Jack Owens

Job: Cascade operator, emergency crew

Age at death: 36

Illness: Rare blood/bone marrow disease

Owen's emergency crew job brought him into some of the most dangerous areas to clean up spills of chemicals and radioactive material. "Some days he'd come home with chemical burns at every orifice," remembers his widow, Norma Rebik. "Later, when his doctor asked what he had been exposed to, he said, `Everything.'" In 1961, at 36, he died of a form of thrombocytopenia, a condition sometimes linked to environmental exposures. "He went from perfectly well to dead in a week," his widow said.

Leon Lindblad

Job: Cascade supervisor

Age at death: 62

Illness: Leukemia

An avowed believer in Paducah's "mission," Lindblad was ambivalent about whether the plant posed risks. "He'd say the radiation levels were not that high," remembers his widow, Virginia, and yet, he always "took his shoes off at the door because he didn't want to bring that stuff inside the house." Lindblad's suspicions multiplied after he became sick with leukemia. He drew up a list of accidents and dates. "If I die, you can sue them," Lindblad explained to his wife, "because they're the ones who did this to me." Virginia never got the chance: On a Friday in 1976, Linblad stashed the list in his desk, never suspecting that he would become gravely ill over the weekend. He never returned to work.

C. Arvil Bean

Job: Process maintenance

Age at death: 64

Illness: Leukemia

Bean's retirement plans included firing up the '49 Cadillac he was restoring and taking his wife on a trip to the Dakotas, where he was once stationed with the Army. Those ambitions faded the day he was diagnosed with acute leukemia at age 55. He replayed in his mind the times he had been exposed to radiation -- like the day he worked 16 hours cleaning up radioactive debris from a 1962 explosion. Despite his illness, Bean clung to his vacation dreams to the end. "Every few days he'd go out there and crank up that old car," daughter Nita said, "even in the snow."

Charles Edward Harris

Job: Machinist

Age at death: 62

Illness: Cancer, multiple organs

For 25 years, Harris worked in the plant's machine shop, grinding down and repairing the nickel-plated pipes and gear used to convert uranium powder to nuclear fuel. Unknown to Harris and most other workers at the time, the metals were contaminated with small amounts of plutonium and neptunium, radioactive elements far more dangerous than ordinary uranium. His son, David, may have been exposed to the same hazards during summer jobs at the plant: College students mowed grass and cleaned up pond sludge in areas now known to be contaminated with the highly radioactive metals. "At the time they told us point-blank there was nothing there but uranium," David said.

Eugene Ragland

Job: Chemical operator

Age at Death: 49

Illness: Lung cancer

The accident and Ragland's death will always be connected, at least in the mind of his widow, Marie. She still remembers his worried voice the night in March 1978 when he called to say he wouldn't be coming home from work. Ragland had been exposed to radiation during a mishap and had been asked to stay overnight for testing. Four months later, a separate medical test found "something wrong" with his blood, she said -- a result that led to the discovery of a rapidly spreading cancer in his lungs and chest. His death on Aug. 4 came so suddenly that Ragland had little time to ponder his illness, or the possible causes. "He always thought he was safe at the plant," Marie said. "They never let him know differently."

CAPTION: When Challie Freeman got sick, Paducah managers claimed he had suffered "no significant" radiation exposure, above, even though years earlier they had restricted his work near uranium, as noted below, because his urine had tested "hot."

CAPTION: The body of Paducah plant worker Joe Harding was exhumed in November for tests. He had feared that the cancer that eventually killed him in 1980 was caused by radiation exposure.