This November 2003 case is two full years before the October 2005 case in Hunan Province, in which a 9-year-old boy had incorrectly been identified as the Chinese mainland's first confirmed case.

The 24-year-old Beijing man who contracted bird flu three years ago died during the outbreak of severe acute respiratory syndrome (SARS). At that time, the man was diagnosed with serious pneumonia without clear a cause. Because testing at that time showed he had not died of SARS, experts kept his sample for further research.

In 2004, when a human case of bird flu was reported in Viet Nam, Chinese scientists began to suspect that the young man who had served in the army had died of bird flu in 2003, according to source from Ministry of Health (MOH).

H5N1 was isolated from the sample.

"So far, the source of infection still remains unknown," said Mao Qun'an, spokesman of MOH. "But this confirmation offers important clues for studying the bird flu outbreak."

There has been no cause of infection determined for the November 2003 case.

The man was a soldier in the PLA, based in Beijing. He did not travel for at least two weeks prior to his illness. It is unknown whether or not he was exposed to any sick poultry or other animals prior to his infection.

There were no outbreaks of H5N1 reported in the Beijing area at the time, according to MOH report submitted to World Health Organization.

Many thought that "many other countries had human cases beginning in late 2003 and early 2004, but not China," said Roy Wadia, spokesman for the Beijing Office of WHO.

"This case now means that China has the first human case in the current H5N1 cycle, which began in late 2003."

"The case in November 2003 shows us once again that whenever H5N1 virus is in the environment, it is possible for humans to get infected and humans do get infected," Wadia said. "It shows us that H5N1 has been in the environment on the Chinese mainland for quite some time now."

China first reported a H5N1 outbreak in poultry in late January 2004. By now, the country has reported 20 human infections and 13 fatalities, including the November 2003 case.

"This latest scientific findings reminds China to improve its disease reporting mechanisms by including the scientific and research institutions in the system," Mao Qun'an said.

The problem is that most of China's human cases have not had a conclusive source of infection, according to Wadia. This is because human cases are found first, without any warning of animal outbreaks in the area.

Agriculture and animal health authorities have had difficulty finding animal H5N1 cases in China, especially if the animals die sporadically in small groups, Wadia said.

The surveillance and reporting systems for both animal and human health need to be strengthened, he added.