While reviewing financial disclosure reports from scientists at the National Institutes of Health, ethics officer Olga Boikess noticed that Dr. P. Trey Sunderland III had not declared any jobs with industry.

In an e-mail sent in March 2000, Boikess told Sunderland: "You did not list any outside positions."

Sunderland, a leading NIH psychiatric researcher, replied: "I do not have any outside positions to note."

In fact, Sunderland had been paid $77,000 in consulting and speaking fees the previous year by Pfizer Inc., now the world's biggest drug company, according to company documents. Between 1998 and 2003, Pfizer paid him $508,050. He did not seek approval to work for Pfizer, and he did not report any of the income to the NIH, as required by agency rules.

Pfizer's payments to Sunderland and his failure to follow the NIH's reporting requirements were described at a congressional subcommittee hearing in June.

Subsequent interviews and government and company documents examined by the Los Angeles Times -- including the e-mail exchange with the ethics officer -- show that Sunderland's paid efforts for Pfizer often overlapped with his NIH role.

Sunderland took the fees from Pfizer at the same time that he led an NIH study of Alzheimer's patients in which the company collaborated.

He also endorsed use of Aricept, Pfizer's drug for Alzheimer's, during a televised presentation at the NIH in 2003. Sunderland did not tell the audience about his affiliation with the company.

Sunderland, 53, is one of the nation's leading researchers on Alzheimer's, the malady that causes dementia in approximately 10% of people over age 65.

He joined the NIH in 1982 after earning an undergraduate degree at Harvard University and a medical degree at George Washington University. As chief of the geriatric psychiatry branch at the agency's National Institute of Mental Health, he has focused on finding ways to detect the disease before a patient develops pronounced symptoms.

Pfizer, along with a corporate partner, Eisai Inc. of Japan, stands to gain billions of dollars in sales from early stage treatment of Alzheimer's. The companies jointly market Aricept, which is approved for treating the symptoms of mild to moderate Alzheimer's. The once-a-day pill generated worldwide sales of $1.6 billion last year, making it the top-selling Alzheimer's drug.

Sunderland also consulted for Eisai from 1999 to 2003, according to information newly provided to the NIH by Sunderland's attorney. Sunderland's income from Eisai was not reflected in documents that the NIH turned over to Congress this year.

Government and company documents show that Sunderland teamed up with Pfizer in both his government and his private roles beginning in 1998. He worked for the company as a paid consultant -- and at the same time led his NIH laboratory in an official research collaboration with Pfizer.

While the NIH allows many forms of moonlighting, the agency forbids its scientists from accepting income from a company that is collaborating with their government laboratory.

The policy seeks to protect the independence of the labs and is consistent with federal law, which prohibits employees from being paid by an outside party for performing government work.

The results of the NIH-Pfizer collaboration, announced in April 2003, underscored the promise of early detection of Alzheimer's. An NIH news release quoted Sunderland, who said such diagnoses "could point to new possibilities for preventive" drugs.

The news release did not mention that Sunderland was a paid consultant to Pfizer.

Investigators at the NIH director's office are assessing whether to refer Sunderland's conduct to the inspector general at the Department of Health and Human Services, documents show.

Sunderland declined to answer questions submitted to him for this article.