It has been three years since Congress ordered the Veterans Administration to investigate whether the use of Agent Orange and other defoliants during the Vietnam war harmed American GIs and entitled them to special compensation.

The VA still hasn't launched that study. It hasn't even decided what scientific protocols or criteria the study will use.

VA critics, mostly Vietnam veterans, claim the agency is stalling because of VA Administrator Robert P. Nimmo's stated concern that the study could result in costs to the government of untold millions if it were to show a direct link between the military's use of toxic chemicals and GI health problems.

Those claims are "politically motivated and absolutely false," said Maurice LeVois, director of the VA's Agent Orange program. "There hasn't been any foot-dragging on our part," added Dr. Vernon N. Houk, chairman of the government's main scientific panel studying Agent Orange.

So far, more than 60,000 Vietnam veterans have complained to the VA of various ailments they blame on Agent Orange, including skin rashes, headaches, nausea and cancer. Some also claim Agent Orange has caused birth defects in their children.

While the VA hasn't officially started the big study, the government is conducting 36 research projects related to Agent Orange and is considering at least 12 new projects, officials said, including one that would compare the health of identical twins, one of whom served in Vietnam. Those projects are proof the government is moving as quickly as possible, officials said.

While such studies may be important, the VA study remains the key to determining whether the government owes special compensation to Vietnam vets. That study, officials admit, is a long way off.

Congress ordered the study after Vietnam veterans claimed the agency was indifferent to their problems, said Ronald Simon, an attorney with the National Veterans Law Center who has been active in the Agent Orange controversy.

Mistrust between Vietnam vets and the VA was so great that President Carter formed an umbrella group--the Interagency Working Group to Study the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants--to oversee Agent Orange studies. The group included officials from the VA and the Defense and Health and Human Services departments.

"It was created to ride herd on the VA and bring it into the 20th Century," Simon said. Jodie Bernstein, the group's first chairman, agrees Carter formed it because of "very serious complaints about the objectivity of the VA." In retrospect, Bernstein said, the VA's problems were not caused by bias, but came about because the agency was unprepared to perform such a study. It did not have doctors on its staff familiar with epidemiological research, she said, nor did it know how to direct such a study.

Neither, apparently, did anyone else. In fact, Bernstein recalled, several government scientists claimed it would be impossible to do an accurate epidemiological survey because no one at the Pentagon seemed to know for sure how many soldiers had come into contact with Agent Orange, and to what degree.

Bernstein and other panel members told Congress that it needed to expand its original order by telling the VA to study the general health of all Vietnam veterans. "If their health was worse than that of people who didn't serve in Vietnam , then we should address why it was that way, and whether the government had a responsibility to do something about that," she said.

While the panel debated whether a study could be done, the VA began seeking a private firm to develop study protocols. When Simon's group saw the VA's plan, it sought a federal court injunction to stop the VA from issuing a contract. The VA plan was "unclear . . . a pile of garbage," Simon said. The judge refused and the VA awarded a contract in May, 1981, for $114,288 to a research team at the University of California at Los Angeles.

By this time, President Reagan had taken office, reorganized the working group and renamed it the Agent Orange Working Group. He gave it Cabinet-level status as a sign of his commitment to problems facing Vietnam vets.

Reagan's actions pleased the veterans, but within weeks they were angry again. A key member of the UCLA team told the California legislature that the dangers posed by Agent Orange had been exaggerated in the media. Nimmo, meanwhile, publicly compared the consequences of Agent Orange to teen-age acne and said Vietnam vets were demanding "preferential coddling." Vietnam vets again charged bias.

In November, UCLA released its first protocol report. The Agent Orange Working Group said it could not even be "classified as a protocol." UCLA said the VA had given it mixed signals about what it wanted.

In March, UCLA returned with a revised protocol that was accepted by the Agent Orange group. That doesn't mean, however, that the study will be underway soon.

The UCLA protocol calls for comparing two groups of Vietnam veterans, one that was exposed to Agent Orange, and one that wasn't, LeVois explained. That's what the VA originally wanted, but last year Congress--worried about the status of the VA study--decided that the VA could expand its mandate if it wanted.

Some VA officials now want to add a third group to the protocol--GIs who were not in Vietnam but who were in the military during the Vietnam era. That would help the VA determine whether just being in Vietnam might have caused GIs health problems.

The working group will decide this fall whether to add the third group, LeVois said. If it does, much of the work done by the UCLA team might have to be discarded and the study would, again, face delays.