new process now undergoing final testing may rid donated blood of virtually all viruses and bacteria, bringing a new level of safety to blood transfusions.

Experts say the process, which uses chemicals that gum up the genetic material of germs, could eliminate the rare cases of AIDS transmission through transfusions that are not already caught by testing blood and screening donors.

Perhaps more important, it could eliminate any other germs, including ones that have not yet been identified.

Had the process been in use 25 years ago, it could have stopped the spread of AIDS through blood transfusions even before scientists had identified the virus that caused the disease.

"No matter what the bug is, you can kill it without even knowing it's there," said Dr. Stephen T. Isaacs, president of Cerus Corporation, the company based in Concord, Calif., that is the leader in developing the technique, called pathogen inactivation.

Cerus hopes to win approval to use its technique on platelets, one component of blood, in the next few months in Europe, where its first clinical trials were begun.

The company also hopes for approval in the United States for platelets by early next year, and for plasma, another component, late in 2003. In addition, Cerus's goal is to win approval to treat red blood cells, which are used in a majority of transfusions, by late 2004 or 2005.

Transfusion experts, however, have mixed views about the technique.

"This technology is tremendously attractive from a blood safety point of view," said Dr. Lawrence Tim Goodnough, a professor at Washington University in St. Louis and director of transfusion services at Barnes-Jewish Hospital.

But he and others also worry that the technique will sharply raise the cost of blood and that the need for the technique is diminishing as new tests are added to further ensure safety.

The Food and Drug Administration has recently approved a new test that reduces the chances of getting infected with H.I.V. or the hepatitis C virus through a transfusion to about 1 in 2 million.

"The technologies are going to be very expensive and the capacity of hospitals and blood centers to support implementation of these technologies financially is going to be a big issue," said Dr. Michael P. Busch, vice president for research at Blood Systems, the second largest collector of blood, behind the American Red Cross.

Cerus and its competitors will also have to prove to the Food and Drug Administration that their process does not introduce new risks that may be greater than the risk of infection that pathogen inactivation would eliminate.

"One of the things we will have to be certain of is that we are not trading a very tiny risk of viral infection for another risk of adding chemicals to the blood," said Dr. Jerry Squires, chief scientific officer of the American Red Cross, which supplies half the nation's blood.

Pathogen inactivation takes advantage of the fact that the parts of the blood that are given in transfusion — red cells to carry oxygen, platelets to help blood clot and plasma for clotting and other purposes — do not contain DNA or RNA, the basic genetic material of life. But bacteria and viruses do. So knocking out DNA or RNA could kill pathogens while, in theory at least, leaving the blood itself unharmed.

Cerus developed a chemical that, when exposed to ultraviolet light, binds to the genetic material. The bonds prevent the two strands of DNA's double helix from unzipping, thereby preventing germs from replicating. RNA, the genetic material in H.I.V., is immobilized in a similar way. Cerus has shown in laboratory studies that its process can eliminate a wide variety of pathogens,

The company was started in 1991 by Dr. Isaacs, a chemist at the nearby University of California at Berkeley who was working on substances that bind to DNA. He met Dr. Laurence M. Corash, a hematologist who had moved to the University of California at San Francisco in 1982, as the AIDS crisis was starting.

"Basically all my hemophiliac patients got H.I.V.," said Dr. Corash, who began searching for a way to clean blood. He is now Cerus's vice president for medical affairs.

Cerus is working with Baxter International, the giant blood products company. Their main competitor is V.I. Technologies, known as Vitex, of Watertown, Mass., which is working with Pall Corporation, a manufacturer of blood filters. Vitex is concentrating on red blood cells and hopes to start the final stage of clinical trials, putting it a little behind Cerus, which is now starting its red cell trials.

Even after pathogen inactivation is approved, experts say, blood will continue to be tested for viruses, and donors will continue to be questioned about their travels and medical conditions, to provide redundant safety measures.

Some experts say they are not convinced that Cerus's technique can completely kill viruses presented in high amounts. Pathogen inactivation will also not be expected to work against prions, which cause mad cow disease and its human equivalent, variant Creutzfeldt-Jakob disease. That is because prions are proteins and do not have DNA.

The biggest hurdle for approval is likely to be safety. If the chemicals bind to the DNA of pathogens, they may also bind to DNA in patients getting the transfusion.

Cerus's platelet system uses a synthetic chemical known as a psoralen. Psoralens are found naturally in citrus fruit, so they are generally considered safe. But there is evidence that some of them can cause cancer. Cerus said its chemical did not cause cancer when tested extensively in mice that had been genetically engineered to develop tumors easily.

Cerus uses a different chemical for red blood cells because light cannot penetrate these cells to activate a psoralen. Vitex uses a chemical it called Inactine, which it has not described publicly. Gambro, a Swedish company, is developing a system that uses vitamin B2, which is already consumed by people. Gambro has not yet started clinical trials.