An AIDS researcher at a laboratory associated with the National Institutes of Health was not following safety rules when he became infected with the concentrated strain of the virus he was handling, federal health officials said yesterday.

NIH officials have known for more than a year that the man was infected with the virus, but only last week could they prove that it came from the laboratory.

"It seems clear that there was a breach of safety rules," said NIH Director James B. Wyngaarden. "We carried out an exhaustive review, and we have concluded that the current level of safety in our laboratories is adequate."

Wyngaarden did not say how the worker erred, but other federal officials said he removed his gloves to take notes in an area where the virus may have leaked from a machine. Officials added that the researcher had skin abrasions on his hands and that the virus that causes acquired immune deficiency syndrome can enter the body through cuts.

The knowledge that a researcher connected with the National Cancer Institute became infected in the lab caused intense anxiety among NIH scientists last week and led to friction among top officials in the health establishment who were unaware of the situation.

The Centers for Disease Control, for example, was never told of the man's infection, even though that federal agency has spent much of the last year revising guidelines to protect health-care workers from the virus.

Researchers at NIH and their families have expressed concern, wondering how a worker who said he followed guidelines carefully could possibly have been infected.

After hours of interviews with the man, who has not developed symptoms of the disease, NIH officials determined that he removed his gloves while working with a viral strain so highly concentrated that it was millions of times more powerful than the one that can be found in humans.

In interviews with health officials, the worker said that leakage from medical instruments containing the virus had occurred and that seals for the laboratory's centrifuge rotors had occasionally failed. The rotors' outer shells could have been contaminated with the virus.

For several years, the NIH has carefully tested hundreds of researchers who work with the virus, and this man is the first laboratory scientist known to have become infected.

"I am certain that the accident had an understandable explanation," said a senior federal health official, who requested anonymity. "But the way the National Cancer Institute handled this investigation is deplorable. They skulked around in the darkness while thousands of people were left ignorant."

The official said that, if lab workers were made aware of the case, they might pay more strict attention to proper safety guidelines.

Wyngaarden said that NCI epidemiologists took months to isolate the virus from the infected worker, failing nine times using a conventional method. Only after Dr. Robert Gallo, the NCI virologist who first discovered that HIV virus causes AIDS, developed a new test method could they be certain the man was infected.

"We just didn't know for certain," Wyngaarden said. "And then it took some time to determine how the man became infected."

AIDS cannot be contracted through casual contact but can pass through abrasions to the blood stream. The virus can only be passed through blood or bodily fluids. The overwhelming majority of AIDS cases are the result of sexual contact or intravenous drug abuse, and the virus has also been passed in blood transfusions and from infected mother to fetus.

The only other documented cases have been caused by infected blood splashing abrasions or accidental swallowing of blood or bodily fluid.

Wyngaarden said yesterday that the mode of the workers' infection appeared similar to cases among health-care workers reported by the CDC earlier this year. At that time, health officials said three women working in hospitals had been infected with the HIV virus after each was accidentally splashed with the blood of infected patients.