Late last year, when many experts were bracing for a resurgence of the SARS virus, Klaus Stohr's thoughts were elsewhere. The head of the World Health Organization's global influenza program was worrying about the bird flu. Trained as a veterinarian, Stohr had long been fascinated by diseases that could make the leap from animals to humans. One such killer bug--the H5N1 avian-flu virus--had killed six of 18 infected people in Hong Kong in 1997. And last fall it was back.

A series of apparently unrelated outbreaks of avian flu had erupted among birds in Vietnam, South Korea and Japan. Then, while attending a medical conference in Okinawa in October, Stohr learned of a possible human death from the flu in China's teeming Guangdong province. Genetic sequencing confirmed that a girl had succumbed to H5N1 in late February 2003--back before SARS sparked a regionwide panic. "Now the H5N1 virus has killed again," Stohr remembers thinking. "A new pandemic could be just a matter of time."

The Guangdong case was never reported in the press, and as recently as last week Chinese officials categorically denied any mainland deaths due to avian flu. What might have been a medical footnote, though, has taken on grim new importance. As of last Friday, 18 people in Vietnam and Thailand had succumbed to the flu, and mass poultry culling and vaccinations were underway in 10 Asian countries. The great fear now is that the virus will mutate into a form that can spread among humans, possibly through genetic mixing with an existing human-flu bug. And China's massive population offers hundreds of millions of chances for H5N1 to do so.

An unpleasant truth is only now coming to light, largely against Beijing's wishes: H5N1 has stalked China for a long time. Before mainland academics and journalists were ordered not to discuss the subject last week, scholars at Huanan Agricultural University in Guangzhou reported that the H5N1 virus had appeared widely in poultry throughout Guangdong, Fujian, Guangxi, Shandong and other provinces as far back as 2001. "I suspect fatal outbreaks such as that in Hong Kong in 1997 have happened many times before in China," says Kennedy Shortridge, an influenza expert formerly of Hong Kong University. "Fifteen years ago children were dying from unexplained illnesses," but awareness of the disease was in short supply back then, he says. Even now many doctors don't suspect bird flu in many cases. "When someone dies of H5N1, the first test just shows death due to influenza A," says bioengineer Henry Niman of Harvard Medical School. "It's a virus you don't find unless you're looking for it."

And that's something Beijing has been reluctant and unprepared to do. After nearly a week of playing hide-and-seek with the media, Chinese officials last week acknowledged "weaknesses and vulnerability in the [country's] animal-disease surveillance system." Since late January, China has owned up to the presence of bird-flu infections in 13 of 31 provinces. But Beijing's stubborn refusal to admit any human deaths due to H5N1 on the mainland--ever--raises questions about how willing authorities are to combat yet another embarrassing crisis.

A year ago, officials didn't admit for months that hundreds of Chinese were dying from SARS. This time the cost of obfuscation could be far higher. John MacKenzie, a scientist who was on the front lines of the WHO's anti-SARS fight, says H5N1 could be "a thousand times" worse than that disease, which killed nearly 800 people worldwide. Recent genetic sequencing has confirmed that the world's worst flu pandemic, in 1918, emerged from a bird-flu virus that also originated in Guangdong. During that scourge, up to 50 million people may have died.

Last week Chinese officials seemed elbow-deep in the battle against H5N1. The village of Kangqiao, outside Shanghai, looked like a war zone. Police blocked every road leading to the community, the site of an avian-flu outbreak. People dressed in bulky white bio-safety suits strode through fields as others hosed down public buses and the occasional passenger car with disinfectant. Yet even the best intentions may not be enough. With 13 billion birds, China is the world's second largest poultry producer and, unlike more developed countries, fully three quarters of mainland poultry farms are small-scale enterprises with less than 100 animals each. The enormous, sprawling size of this network makes early virus detection little better than guesswork.

No one needs to tell that to Xue Xuelong. The 40-year-old farmer initially didn't even bother to report that nearly all his 1,700 chickens had dropped dead in mid-January. After seeing a televised news report on the flu, Xue told local officials, who had his remaining birds slaughtered. But last week authorities told Xue his birds didn't have avian flu after all. He still doesn't know what killed his flock.

While local officials may have decided it's better to be safe than sorry, it doesn't appear their superiors have learned the same lesson. International health experts report that the Ministry of Agriculture--which is responsible for animal diseases--has been "foot-dragging." Stohr says a top priority now is to gather virus samples from all affected countries so that an appropriate vaccine can be developed. But China hasn't been included in this effort "because we haven't received anything from them," he says. If officials admitted the disease had spread to humans, greater responsibility for handling the outbreak might be handed to the Ministry of Health, which many experts credit with reforming itself after the SARS debacle. But officials still haven't addressed the mysterious case of the girl who died from H5N1 last February. Last month the WHO's China office submitted a report detailing H5N1's role in the death to Beijing authorities, after they repeated that there had never been any human cases of H5N1 in Guangdong. It could be China's most dangerous denial yet.