ARMY MAJ. Charlie Hubbs used to drink Agent Orange during the Vietnam War to convince reporters it was harmless.

"It tasted terrible," recalls Hubbs, now a 52-year-old resident of Camp Springs, Md., "but I haven't changed my mind." He still believes the herbicide, used to clear jungle and kill crops, hasn't damaged anyone's health, including his own.

Michael Ryan never drank Agent Orange, but the Long Island policeman was exposed to it in Vietnam. He believes it caused his 12-year-old daughter, Kerrie, to be born with 18 birth defects, including missing bones, twisted limbs, a hole in her heart and no rectum.

Their conflicting views are not only at the core of the Agent Orange dispute but central to the Environmental Protection Agency's $33 million political decision last week to buy out the entire town of Times Beach, Mo.

The chemical that made Times Beach unfit for its 2,400 inhabitants is dioxin. Dioxin -- found in Agent Orange at levels more than six times greater than that detected in Times Beach -- is what many Vietnam vets charge has caused them everything from headaches to cancer to birth defects in their children. So far 16,102 vets have filed disability claims.

Moreover, some estimates have claim that as many as 165 million Americans may have been exposed to dioxins in the years they have been sprayed on forests, crops and roadsides and home gardens. The herbicides have been used to control weeds at the Washington Monument and on the White House lawn.

In yet another twist, EPA still lets chemical manufacturers sell dioxin- ridden herbicides for use on rangeland and rice -- 1.3 million gallons were sprayed on rangeland last year. In fact, EPA has moved recently to ease restrictions on the dioxin contaminated herbicide, 2,4,5-T.

All these contradictions simply underscore the key point about the government's conflicting stands on dioxin: political decisions are being made more on headline appeal than on scientific evidence. The bulk of reputable science indicates simply that no one knows whether dioxin is toxic to humans and if so, how much.

Strange as it may sound after a week of reports describing dioxin as a threat to humankind, scientists are deeply divided over whether exposure to dioxin causes health problems in humans.

Consider these comments by two experts on dioxin: Drs. Samuel Epstein, a cancer researcher at the University of Illinois, and Matthew Meselson of Havard, considered a pioneer in dioxin research.

Epstein: "The evidence is overwhelming that dioxin is carcinogenic (causes cancer) in humans."

Meselson: "Anyone who says that it is a clear and present danger cannot prove it, and anyone who asserts that it is not a danger cannot prove that either. We simply don't know."

Dioxin is formed as an unwanted byproduct when "phenoxy" herbicides are made. Just how much dioxin is formed depends upon how well the manufactuer controls the temperature and pressure when making a herbicide.

In 1979, the President's Council on Environmental Quality announced that dioxin had caused cancer in test animals when 2.2 parts per billion of their normal diet consisted of dioxin. It caused birth defects in monkeys at doses of 2.5 parts per trillion. The council called dioxin "one of the most toxic substances ever studied."

The Centers for Disease Control in Atlanta now says dioxin at levels higher than one part per billion may be dangerous to humans.

But dioxin effects animals differently, making some scientists skeptical of comparisons with humans. For instance, the dosage needed to kill a hamster is 5,000 times greater than that needed to kill a guinea pig, according to Dr. Barclay M. Shepard, the VA's expert on Agent Orange.

Epstein claims other chemicals have been banned after causing cancer at much lower doses than dioxin. The government is stalling to avoid paying compensation, he charges.

Most of what is known about the effect of dioxin on humans is based on studies of industrial accidents, primarily one in 1949 at a chemical plant in Nitro, W. Va., and another in 1976 in Seveso, Italy.

The Nitro accident occurred when a value popped, spraying fumes inside a building. After that mishap 121 workers developed chloracne, a skin problem caused by dioxin. Studies have documented numerous short-term effects after the accident, including nausea, headaches, blood disorders and impaired nerve function. But the symptoms diminished over a period of time and eventually disappeared. Recent studies show the death rate among the 121 men actually was lower than normal. There also were no signs that the men had developed higher rates of cancer.

At Seveso, an explosion spewed an estimated 10 ounces of dioxin over 700 acres. Within six days, 11 children had been hospitalized with skin inflammation and 175 cases of chloracne had been diagnosed. The government evacuated 212 families from homes near the plant after the dioxin level was found to be 50 parts per million. Two months later more than 2,000 small animals had died, mostly rabbits. Another 78,000 animals had to be slaughted to prevent dioxin from entering the food chain.

Officials reported a signifigant increase in birth defects in Seveso and a higher rate of spontaneous abortions. But both rates were lower than the country's national average.

In January, Vietnam invited 21 nations to send scientists there to review 60 herbicide studies by the Vietnamese. Fourteen scientists from U.S. universities attended, but they concluded that none of the reports was scientifically valid.

The only U.S. study linking a herbicide with dioxin to human health problems was issued in 1979 by the EPA. It was prompted by Bonnie Hill of Alsea, Ore., after she and seven other women in the Alsea area suffered miscarriages several weeks after 2,4,5-T was sprayed near their homes. The study said the miscarriage rate "appears to be correlated to the amount of 2,4,5-T sprayed," prompting EPA to limit 2,4,5-T use to rangeland and rice.

More than 18 critiques of the Alsea II study have been published in scientific journals. In 1979, the British medical journal The Lancet described the sentiment among scientists like this: "Independent statisticians have been unable to find any evidence in the (Alsea II) data of a link between abortion and 2,4,5-T."

In 1980, the VA hired a consulting firm to review some 1,200 scientific articles about phenoxy herbicides. The study found "no evidence" that the chemicals pose "a potent carcinogenic risk" to humans. It also found no evidence that dioxin caused reproductive damage in humans.

Last October, the American Medical Association issued its 2,4,5-T and dioxin studies. It said "there is still little substantive evidence for the many claims that have been made against these compounds."

Dow has even given volunteer employes doses of 2,4,5-T to see how long it took to go through their bodies.

The VA is involved in 50 studies alone this year of Agent Orange, at a cost that will exceed $100 million. EPA will spend another $495,000 to study residents of Times Beach and other areas exposed to dioxin.

Suspicion remains, of course, about the VA's objectivity. Congress last year forced VA to surrend its most important study to the CDC because it said the VA had lost credibility with Vietnam veterans. The scientific dispute, however, may now be mute. "We're simply beyond that now," says Frank McCarthy, president of Agent Orange Victims International, a privately funded self-help group.

How can the government pay $33 million to 2,400 residents of Times Beach and not feel responsible for "the healthy young men it sent to Vietnam who came back sick and diseased?" he asks