A small Rockville biotechnology company has created an antidote to anthrax that protects animals against extremely high doses of the biological warfare agent and, based on testing done so far, may do the same for people.

The company, Human Genome Sciences Inc., plans to announce this morning that it is preparing to launch human tests of the drug in a matter of weeks. If the tests go smoothly, the drug may well enter the nation's arsenal against bioterrorism by next year.

Scientists would still lack firm information about when and how to use the drug, given that only limited human tests are possible. Still, the availability of the drug would represent one of the first concrete results of a pledge that American scientists made, after Sept. 11, 2001, and the ensuing anthrax attacks, to deploy their skills in defense of the nation.

The project has unfolded in complete secrecy but with remarkable speed. In 16 months of intensive effort -- one team leader padded around the laboratory in his pajamas over a holiday break to tend to critical experiments -- Human Genome Sciences has reached a point in developing a new drug that often takes five years or more.

The company recently detailed its progress for key government leaders. Anthony S. Fauci, director of the National Institute of Allergy and Infectious Diseases and a prime leader in the government's anti-terror efforts, called data from animal tests "impressive" and said the drug could become an important tool in the country's effort to build defenses against biological terrorism.

Other laboratories are known to be working on approaches similar to the one pursued by Human Genome Sciences, but partly because the company spent its own money, and did not wait for government research grants, it appears to be ahead. Interviews with two companies pursuing similar antidotes suggest that they are many months, at best, from starting the sort of live-anthrax animal tests that Human Genome Sciences is finishing.

By happenstance, the company is completing the experiments just as President Bush and his allies in Congress are preparing to seek approval of Project BioShield, a $6 billion initiative Bush outlined in his State of the Union message that would give the government standing authority to buy and stockpile drugs, vaccines and the like to counter terrorist threats.

When the people at Human Genome Sciences briefed Fauci on their recent results, "I thought, 'My God, this could be the poster child for BioShield,' " Fauci said.

The goal of the tests the company is about to start will be to determine whether the drug is safe to give to people, and that can be established in a few months. Tests of how well the drug would work in those exposed to anthrax are not possible -- barring deliberate attack, the disease occurs very rarely and intentionally exposing people to the spores is unethical.

In tests involving monkeys and rabbits, which are the best available proxies for people in this type of research, some levels of the drug, administered shortly before or shortly after exposure to high doses of anthrax spores, protected all the animals from death, said Craig A. Rosen, president of research and development at Human Genome Sciences.

Moreover, the drug appeared to last a long time in the animals' bodies and might retain its effectiveness in people for as long as a month, Rosen said. Whether it could, however, achieve anything like complete effectiveness if administered to people days after exposure to anthrax is likely to remain uncertain.

The Food and Drug Administration has declared its willingness to approve anti-bioterrorism drugs based on proof of effectiveness in animals alone. Tests involving monkeys may be regarded as particularly significant, since the behavior of drugs in monkeys and humans is often similar.

Theory suggests that the drug could be useful as an adjunct to antibiotic treatment for people exposed to anthrax, or as a preventive measure for soldiers who have not yet completed the arduous vaccinations required for immunity to anthrax. The drug might even serve as a last resort for sick people in whom other treatments have failed.

And, perhaps most important, a drug of this type might be the only recourse if a terrorist created a strain of anthrax resistant to the standard antibiotics used to treat the illness -- an easy enough feat for anyone capable of growing anthrax in the first place.

Precise human data about the drug's usefulness in all these situations would probably be possible only after another anthrax attack, which means the government may have to make decisions about buying the product, and others under development, amid considerable uncertainty.

William A. Haseltine, chairman and chief executive of Human Genome Sciences, said in this case his company -- one of the best-financed small biotechnology companies in the country, with $1.5 billion in the bank -- was able to pay for the experiments without raising new funds. But for other companies and future projects, he said, passage of a measure like Bush's BioShield plan will be critical to drawing investor capital.

"Without BioShield, it is difficult to imagine how we or other companies could create these projects in a timely fashion," he said.

The new drug, which the company has named Abthrax, is a genetically engineered antibody given by injection. Antibodies are immune-system proteins that attack and help to disarm invading organisms. Only in recent years have scientists mastered the art of making them artificially, resulting in some remarkably effective drugs.

The project began soon after the anthrax attacks in 2001, as experts at Human Genome Sciences discussed the potential usefulness of an antibody that would attack a toxin produced by the anthrax bacterium, Bacillus anthracis. Neutralizing this toxin could have stunning effects, they theorized.

On the first Pearl Harbor Day after the Sept. 11 attacks, Rosen called aside a young microbiologist named Michael Laird and gave him an unusual assignment: Not only was he going to make an antibody, but to save time and paperwork in obtaining the necessary gene for the anthrax toxin, Laird and a small team of assistants would create it artificially, by painstakingly stringing together tens of thousands of units of genetic information.

Laird said he's normally a workaholic anyway, skipping a night's sleep every week, but the demands of this project were extraordinary. "Christmas Day I was at home for three hours out of the 24," said Laird, the father of two young children. "I was running around here in flannel pajamas and a baseball cap in the middle of the night."

Through arduous months of work, the company said nothing publicly, mindful that premature promises on such an important issue as anthrax defense could cost it credibility. Eventually, Laird and his team managed to produce huge amounts of anthrax toxin, create the antibody, make large quantities of it and begin testing the drug in animals.

Laird recalls learning results of some of the first rat experiments, when a control group injected with the toxin alone died, while animals given high doses of the antibody lived. It was hard proof the company was on the right track.

"I was exuberant," Laird said. "I went and ran 20 miles." He called his wife after jogging, but with the project still under wraps, he couldn't tell her exactly why he was so happy.