After infecting more than 8,400 people and killing at least 812, the SARS epidemic appears to have been beaten back, with the virus no longer finding new victims anywhere in the world.

If no new cases of SARS are reported in Taiwan by today, the World Health Organization will remove the island from the unenviable list of places where the virus is being transmitted from person to person. Taiwan remained alone on the list after battered Toronto came off three days ago.

The move means no new cases of severe acute respiratory syndrome have been reported on the island for 20 days -- twice the incubation period for the virus.

Around the world, health authorities are cautiously relieved that the "chain of transmission" appears to have been broken for a deadly disease that several months ago appeared to be spiraling out of control.

"This is a great achievement for public health," said David L. Heymann, executive director of WHO's communicable diseases program.

But any sense of triumph on the part of Heymann and others is decidedly subdued. Although the disease has dropped out of the headlines, health officials know all too well that much about the frightening new virus remains mysterious. It could reemerge anytime, perhaps in an even more deadly or contagious form, they said. SARS has also served as a dramatic reminder that dangerous new microbes can emerge without warning, threatening a devastating pandemic.

"It's satisfying that we've been able to contain it, but that could be our worst enemy," Heymann said in a telephone interview. "We can't let down our guard."

Heymann noted that dozens of people are still sick from SARS, so there could be more deaths from the disease. "There are some who are quite seriously ill," he said.

And, Heymann added, he cannot rule out the possibility that additional cases could emerge. A second outbreak flared in Toronto, for example, after the first spate of infections appeared to have been controlled.

Most important, Heymann and others said, viral diseases, particularly respiratory infections -- even familiar ones -- are notoriously unpredictable. That makes it impossible to know what the future may hold.

"Let's be optimistic, but cautious," said Robert G. Webster, a leading virologist based at the St. Jude Children's Research Hospital in Memphis.

Because of the continuing uncertainty and possibility of grave danger, researchers are urgently working to answer the most pressing questions about the SARS virus.

An intense effort is underway in southern China to test animals to determine whether catlike mammals known as civets are the primary hiding place for the virus in nature, or whether other species also play a role. Knowing the answer will be key in preventing the virus from jumping again from animals to humans.

"How many animals are infected? How often do they have contact with humans? How much virus is excreted by these animals?" said Klaus Stohr, WHO's top SARS scientist, from China, where he has been working with local authorities to set up a research agenda. "If there is an animal reservoir, then we have to realize that the virus might come back."

Researchers are also working to develop treatments and a vaccine, but the most urgent work is focused on finding a faster and more accurate way to test for the virus. Because the flu-like symptoms of SARS are so similar to other diseases, another outbreak could quickly overwhelm hospitals and public health departments if there is no way to identify the SARS patients.

Another key question is whether the virus is still spreading but causing milder disease or atypical symptoms.

"It may still be present and transmitting in very low levels, and causing disease that isn't picked up by our surveillance system," Heymann said.

Doctors in Hong Kong recently recommended that WHO broaden its definition of "suspected cases" after determining that there were infected people in that city who would be missed by the current definition.

It is also possible that some people can become infected and spread the virus without developing symptoms. If that is occurring, most officials believe it is rare, because they have been able to trace nearly all reported cases back to their source.

Still, said one well-known specialist, "it would be unusual to have an infectious disease where you don't have some asymptomatic infections."

"The million-dollar question is, 'Do these people pose a risk of infection?' " added Michael T. Osterholm of the University of Minnesota. "We just don't know."

Many respiratory viruses, such as the flu and colds, follow a seasonal pattern, with outbreaks peaking in the winter and subsiding in the summer. Some experts believe that natural cycle may have played a role, perhaps a major role, in causing the SARS epidemic to subside, and that it will likely come back in the winter.

"The efforts to control SARS were among the very best in public health in the last 50 years. But I'm not so sure it had anything to do with the elimination. I think seasonality had a huge role," Osterholm said. "Just as the flu goes away, I think SARS may have done the same thing. I believe early next winter in the Northern Hemisphere we will see a return of SARS."

That would be especially problematic, because it would come at the same time as there are flare-ups of other diseases with similar symptoms, such as influenza, making it difficult to distinguish SARS from those illnesses.

"It will be very difficult until we have a good diagnostic test," Heymann said.

Researchers also want to determine why some people can be highly infectious.

Although most SARS victims infect only two or three other people, the epidemic has been punctuated by the appearance of "superspreaders," who inexplicably infect large numbers of people. In one incident in Toronto, for example, a woman infected more than a dozen patients, staff members and family members in a hospital within 45 minutes, including some who do not remember any direct contact with her. In the most extreme cases, a single patient has spread the disease to perhaps 100 others.

Mutations in the virus's genes have been detected, but researchers are watching carefully for any signs that the changes are making the virus more deadly or contagious, as well as more difficult to fend off with a vaccine.

SARS emerged in November in the southern Chinese province of Guangdong. The Chinese government tried to conceal the outbreak for several months, which allowed the virus to spread around the world.

After major outbreaks erupted in Hanoi, Toronto, Singapore and Toronto, WHO declared an international health emergency on March 15. It was the first deadly respiratory virus that could spread directly from one person to another to emerge in decades, and officials feared it could mushroom into a global pandemic.

The sense of alarm heightened as the number of cases mounted, and the death rate steadily rose, reaching 15 percent overall. Among the elderly, about half died.

Because there are no tests, treatments or vaccine, the only way to contain the virus was to resort to primitive public health measures that had not been widely used for decades -- identifying victims based on their symptoms and isolating them as quickly as possible to limit the number of people who could be infected. The disease was spread primarily by small droplets spewed out when infected people sneezed or coughed.

An unprecedented international collaboration identified the cause of the disease in record time -- a virus in the corona family. Previous coronaviruses caused only colds in people, though they were known to cause serious respiratory and gastrointestinal illnesses in animals.

Scientists eventually identified a virtually identical virus in civets, a delicacy in southern China. It was later also found in raccoon dogs.

"Right now, we are about to watch the history of this epidemic take a right or left turn at the fork," Osterholm said. "We don't know which way it's going to go."

A woman reads next to a SARS-prevention billboard at a train station in Taipei, Taiwan.