"We are forced to conclude that there is no scientific basis to believe that Peptide T is effective."

With that remark during a plenary session of the Third International Conference on AIDS here yesterday, Dr. William Haseltine of Harvard University made public a controversy brewing for months just below the surface of scientific and professional civility.

"It has been very upsetting," said Dr. Candace Pert, a neuroscientist at the National Institute of Mental Health and discoverer of Peptide T.

The debate appears to be as much about the politics of science and the intense competition among groups within the AIDS research community as about science. Research on acquired immune deficiency sydrome has produced a number of controversies, including the effectiveness of another drug, ribavirin.

Peptide T, a short stretch of protein building-blocks, purportedly mimics part of the virus that helps it infect white blood cells. If the fit is correct, Peptide T should prevent infection, just as one key in a lock blocks entry of another.

Pert said Peptide T fits the cellular lock. Haseltine disagrees.

"Peptide T does not work. Nothing. Nada," he said in an interview. He said he believes that the peptide fails to block the virus because it does not fit any sensitive part of the target molecule on the cell's surface.

Haseltine added that several laboratories around the world have been unable to repeat Pert's experiments. Several other scientists, who asked not to be named, also said they have questions about the Peptide T data.

Yesterday, however, Dr. Elaine Kinney-Thomas from Oncogen, a Seattle biotech company, reported that Peptide T blocked AIDS virus infection in laboratory-grown cells.

The intensity of the debate arises, in part, because tests of Peptide T's ability to slow the AIDS virus have been conducted on humans.

Dr. Lennart Wetterberg of the Karolinska Institute in Stockholm has tested it in four patients and initially came to present results of his study yesterday. Instead, he was asked to present a more general talk about the neurological impact of AIDS.

Dr. George Galasso, an assistant director of the National Institutes of Health and the conference chairman, confirmed that Wetterberg was asked to not concentrate on Peptide T. "All plenary speakers are supposed to talk about a broad spectrum," he said.

Other tests on humans are being planned by Pert's group, which has received Food and Drug Administration approval.

Wetterberg reported that three of his four patients, all near death when treatment began, appeared to do well. The fourth died.

Those results led Wetterberg, who said he does not know whether the drug works, to begin a controlled study with 36 patients, half of whom will receive the drug.

"That," he said, "is the only way to say whether Peptide {T} is effective."