Dr. Thomas J. Walsh, a physician-researcher for the National Institutes of Health, has added to his income and bolstered his lab with checks from drug companies, according to interviews and government records.

The NIH would not comment on Walsh's relationship with industry, including whether he had obtained permission for outside work or had reported the income.

Long-standing agency rules have generally prohibited scientists from accepting fees from companies in cases where private and government work overlap. Last year the NIH barred its scientists from taking any consulting payments directly from industry.

Collaborations with industry that exclude the payment of personal fees to NIH employees, however, remain permitted and occur regularly. Both old and new rules have required scientists to get permission to do outside work, and to report any compensation.

In 2004, the most recent year for which complete summaries were obtained, Walsh received $153,627 in federal compensation, including $30,242 in nontaxable housing and living allowances. His unabridged annual financial reports, including details of any outside income, are confidential, as are those of most senior NIH researchers.

In April, Walsh said in a statement to The Times that collaborating with industry was integral to his government work. He did not discuss specifics of his ties to companies for this article.

Nor has Walsh always disclosed such ties when writing for a medical journal.

An editorial by Walsh, published in December 2002 by the New England Journal of Medicine, referred to an antifungal drug made by Merck & Co. -- but did not mention his receipt of fee income from the company. The journal's conflict-of-interest policy requires authors to acknowledge such income.

After inquiries from The Times, the journal asked Walsh to submit a revised disclosure statement. A spokeswoman for the journal, Karen Pedersen, said Walsh had "failed to disclose" the income from Merck when he submitted the editorial. On March 16, the journal published a correction, saying the editorial "should have included the fact that Dr. Walsh received an honorarium from Merck."

Walsh told The Times that his lack of disclosure was inadvertent.

Walsh also has consulted with Pfizer, said Mariann Caprino, a spokeswoman for the company. She said Walsh was paid from 2001 to 2005 for participating in private meetings at Pfizer "to discuss clinical trial designs."

"We did compensate him for his time," Caprino said in an interview.

When Pfizer brought its antifungal drug before an FDA advisory committee in October 2001, a company executive identified Walsh as part of its "sponsor section." The FDA approved the drug, Vfend, about seven months after the advisory committee meeting.

"He attended the advisory committee in the capacity of an expert in this area," Caprino said, adding: "His relationship was fully disclosed by us at that meeting."

Walsh said that he was not paid by Pfizer to attend the meeting. He did not respond to questions seeking details about his compensation from Pfizer.

Caprino declined to state how much Pfizer had paid Walsh over the years, but said: "He received a standard per-diem rate. Basically, that compensates him for time out of the office."

Merck and another company, Fujisawa USA Inc., in recent years also identified Walsh both as a company participant and as an NIH official when the firms presented new antifungal drugs at FDA advisory committee hearings.

A spokesman for Merck, Christopher Loder, said the company had paid Walsh personal fees of $2,000 in 1999 and $1,000 in 2001 for participating in two "Merck meetings." Loder said Merck did not pay Walsh fees to develop its new antifungal drug, Cancidas.

As for Fujisawa, in 1997 Walsh assisted the company's presentation of its first antifungal drug to an FDA advisory committee. And on March 28, 2003, Walsh accompanied Fujisawa executives when they met privately with FDA staff to discuss the company's latest antifungal, Mycamine. A summary of the meeting prepared by the FDA identified Walsh as a consultant.