A psychiatrist has been treating AIDS victims by injecting them with blood containing the live HIV virus in an experiment that has met with serious doubts from leading researchers and ethicists, it was reported Saturday.

The doctor, Michael J. Scolaro of Los Angeles, said that all of the patients had volunteered and had "already failed both approved and experimental therapies" and were facing death.The experimental therapy is designed to fight AIDS by introducing a reputedly non-disease producing strain of HIV, or human immunodeficiency virus, into the body so that it can compete with the patient's own virulent strains.

The virus was obtained from an individual who has been apparently infected with HIV for more than 10 years but who has remained free of symptoms, the Los Angeles Times said.

But prominent AIDS researchers and medical ethics experts have raised serious questions about the scientific validity of the theory on which the study is based, as well as the safety of the experiment.

There is no evidence that the feasibility of the experiment has been demonstrated in laboratory tests and animal experiments before it was tried in humans.

"At first sight, the whole thing looks screwball," Albert Jonsen, chairman of the department of medical history and ethics at the University of Washington School of Medicine in Seattle, told the newspaper.

For his part, Scolaro, who declined to be interviewed, said in a statement that the "preliminary findings" represented a "potential therapeutic modality, which thus far, has been overlooked in AIDS research."

The experiment, the Times said, began last sumer and is continuing. It involves 11 severely ill AIDS victims who were getting worse despite treatment with AZT and other drugs.

Each of the patients received two inoculations of less than a teaspoon of whole blood from the apparently healthy HIV-infected individual.

After 41 weeks of follow-up, four patients have improved, three had mixed responses, three regressed and one died, according to a press release issued when a little noticed brief account of the research was published March 23 in The Lancet, a British medical journal.

All 10 survivors had lab evidence of increasing immunity, according to the account. Those who improved had increased amounts of a protein found in the reputedly harmless AIDS virus strain. No new symptoms were noted, aside from short-term muscle aches, fever and diarrhea.

AIDS experts said it was impossible to draw any conclusions one way or another from the data.