In a unanimous vote in Atlanta, the Advisory Committee on Immunization Practices said the high probability of dire vaccine side effects, combined with a low possibility of bioterrorist attack, made the vaccine too risky for widespread use.

Instead, the 15-member panel proposed that state governments and bioterrorism authorities designate two smaller groups to receive the vaccine: teams of investigators who would track down potential cases of smallpox, and staff at a small number of hospitals that would be designated to receive infected patients. Overall, that could total 10,000 to 20,000 Americans, the scientists said.

Potential recipients would be screened to ensure they and their families have no identifiable risk of vaccine side effects, which range from mild scarring to rare cases of neurological complications and death.

"The vaccination will always be voluntary," said Dr. John Modlin, the committee chairman.

The recommendation now goes to the acting director of the Centers for Disease Control and Prevention, who will submit it to Health and Human Services Secretary Tommy Thompson.

"We plan to move as expeditiously as possible," Thompson said late Thursday. "While the threat of a smallpox attack is believed to be low, it is indeed possible, and unfortunately we now live in a world where we need to be prepared for the possibility."

It could be three to six months before HHS allows vaccination of the state teams to begin, CDC staff said. Nevertheless, state governments are expected to press ahead with their bioterrorism response planning, which is a precondition for receiving some of the $1.1 billion Congress appropriated for improving local bioterror preparations.

State participation is key, said Dr. Julie Gerberding, acting deputy director of the CDC: "We learned last fall that we can get [medical supplies] anywhere in the United States in 12 to 24 hours. The limiting issue is not getting the vaccine there; it is, once it is there, how does it get into the arms of the people who need to receive it."

State governments will have to nominate at least one bioterror response team -- probably 10 to 12 doctors and nurses, epidemiologists, lab workers and law enforcement officers -- and designate at least one hospital where smallpox patients would be confined and treated.

In each of the hospitals, one day shift and one night shift -- from doctors to technicians to maintenance and laundry staff -- probably will need to be vaccinated, said Dr. Ed Thompson, state health director for Mississippi.

Meanwhile, Gerberding said, the CDC will create a board of experts who will coach state governments through naming the teams and planning the vaccinations. In Georgia, officials were relieved that widespread vaccination was not approved.

"We are very happy they are not recommending wholesale immunization of the population," state epidemiologist Dr. Paul Blake said. "This is a very dangerous vaccine. It is not something you would like to see in really widespread use unless the need was really desperate."

Georgia has been working on bioterrorism preparation since the 1996 Olympics, but has not named a smallpox investigation team or smallpox hospital, Blake said.

"We know that other states have had some resistance from their hospitals, so we have not yet talked formally with any hospitals in Georgia about that," he said.

The panel's recommendation followed public meetings in five cities to gather comment from scientists, emergency and medical workers, and the public. Much of the testimony was passionately opinionated.

In a meeting in Atlanta on Thursday, for instance, noted vaccine researcher Dr. Stanley Plotkin criticized the basis of the vaccination plan: a hypothetical scenario in which the virus is brought into the country by a few infected individuals and has not been genetically modified to be more lethal.

"The assumption you have made is that the smallpox we will face is the smallpox we know," Plotkin said. "I think there is a danger of our using the tactics of the last war to fight the current one."

Dr. D.A. Henderson, who led the global effort to eradicate smallpox in the 1970s and is a special adviser to Secretary Thompson, questioned whether the plan extended enough protection even to medical workers.

"It concerns me that we are going to have just so many hospitals designated as smallpox hospitals," Henderson said. "If we get into an epidemic situation, we will end up with a fair number of people and hospitals involved whether they want to be or not."

Smallpox was eradicated in 1979, and since then has existed only in laboratory freezers at the CDC and in Russia.

There have been no U.S. smallpox cases since the 1940s. No smallpox vaccinations have been given in the country since the 1980s, except to some CDC staff.

Smallpox planning is complicated by uncertainty over the means of preventing the disease. The vaccine used in the 1970s has been held by the government since then. But because part of the stockpile became too old to be used and had to be remanufactured, the vaccine was reclassified by the Food and Drug Administration from a licensed drug to an "investigational" one. . Now, counseling sessions and sheaves of paperwork are required before it can be given to anyone.

In addition, the drug used to treat serious side effects -- made from the blood of those who have received the vaccine -- is in critically short supply.

A newer vaccine is being prepared. It has been made with more modern laboratory methods and may cause fewer serious reactions, though that has not been proven. But the new vaccine will not be fully licensed for another 18 months, and will first have to undergo clinical trials in adults. It may not be tested in children, a prospect that clearly alarmed several panel members.

"If there were an actual smallpox case, we would be vaccinating children in the absence of available data about the vaccine," said Dr. Paul Offit, a committee member from the Children's Hospital of Pennsylvania.

Explaining their decision was on the minds of panel members Thursday.