WASHINGTON — The National Institutes of Health has banned its scientists from working as consultants for pharmaceutical, biotechnology and similar businesses, a step that officials hope will end controversy over paid consulting arrangements by some of its doctors.

The new rules, issued by the Department of Health and Human Services and the Office of Government Ethics, cap more than a year of investigations into allegations of conflict of interest involving NIH scientists and administrators.

"Nothing is more important for NIH than preserving the public's trust," Dr. Elias Zerhouni, NIH director, said yesterday.

Under the system, employees may not be paid to consult, speak or write for any drug or biotech company, medical-device maker, health provider or insurer, or trade or professional group. Thousands of employees will have to divest themselves of all stocks in drug, biotech and related companies, with the exception of diversified mutual funds.

Even the lowest-ranking employees with no control over purse strings or policies will be subject to new limits of $15,000 on health-related stock holdings.

NIH scientists will be allowed, however, to teach courses and give lectures related to their work, as well as write articles and textbooks.

"Clearly we do not want to impair scientific interchange," Zerhouni said at a briefing.

Zerhouni also said that the NIH last month established rules requiring researchers who receive royalties from experimental treatments they helped develop to disclose those royalties to patients.

The new ethics rules will take effect when they are published, probably within a day or two.

The changes codify the reversal of a trend toward liberalized links between NIH researchers and drug and biotech companies that began in 1995, when government ethics officers criticized the NIH for having restrictions on outside consulting that were stricter than those at other agencies. Harold Varmus, NIH director at the time, responded by loosening the rules.

In December 2003 the Los Angeles Times reported that some NIH scientists had arranged lucrative side deals that, though apparently legal, raised concerns about conflict of interest.

In perhaps the most egregious example, a researcher at the National Institute of Mental Health was paid $517,000 in fees, honoraria and expenses as a consultant for the drug giant Pfizer over five years without disclosing that income to the agency, according to congressional investigators.