When the State Department asked him to appear in an ad campaign presenting America as a land of tolerance to the world's Muslims, Dr. Elias A. Zerhouni didn't have to think long. He had a story to tell.

Twenty-seven years ago, the Algerian-born Zerhouni came to this country with $369 in his pocket. Just out of medical school, he trained at the prestigious Johns Hopkins School of Medicine and rose through the ranks there to become radiology chairman and finally vice dean for research.

Now he is six months into his tenure as director of the National Institutes of Health, steering the nation's research agenda on issues such as stem cells, cancer, heart disease and biodefense.

"I don't think there is any other country in the world where different people from different countries are as accepted and welcomed as members of a society," Zerhouni, 51, said in ads that also featured a Libyan-born baker, a Lebanese-born teacher and a Muslim fire department medic.

Recently he sat down to talk about the past six months and what lies ahead.

He has made his feelings clear about being an American - "I think I represent what's good about America, and I'm proud of saying it" - but many people are less clear about where he stands on one of the most ethically charged issues in biomedical research: harvesting stem cells from human embryos to cure the sick.

At Hopkins, Zerhouni was instrumental in linking an anonymous donor of $58.5 million with a new institute for cell engineering that will focus primarily on stem cells. Now, at NIH, he must pull off a delicate balancing act - advancing the science while heeding the president's policy that allows federal funds for research on existing stem cell lines but not for the creation of new ones.

The policy reflects President Bush's ethical qualms about destroying embryos even to cure the sick, though he supports letting scientists strive to understand stem cells and tap their potential.

Zerhouni said the Bush administration didn't ask him to support the policy before it offered him the nomination, bristling at any suggestion of a litmus test. But he defends the policy, saying it allows scientists to come to fundamental understandings about stem cells before rushing into experimental therapies that could prove dangerous.

"This field of science needs to walk before it can run," Zerhouni said during an interview at his office on the 300-acre NIH campus in Bethesda. "It's going to be a long road, and we need to know how stem cells really do what they do, make sure they don't turn malignant, how is it that we can control this."

Even while helping to shape the cell-engineering institute at Hopkins, Zerhouni said, he was troubled by scientists at some institutions whose exaggerated forecasts created false hopes that cures were close at hand.

"As a scientist, I knew that many of the claims were hype, not substance. So to me, what was really important was to free up federal funding, to explore the field."

But many scientists complain that the limitation has slowed the pace of research. Although the NIH first said there were 76 stem cell lines in the United States and abroad, researchers quickly argued that only a handful were viable and that many of those were off limits because of commercial barriers.

Dr. Curt Civin, a Hopkins oncologist, told a Senate subcommittee in September that he had spent a lot of time trying to buy cells and had come up empty. Some approved stem cell lines were technically flawed, and at least one nation - India - had barred exports of them.

"Embryonic stem cell research is crawling like a caterpillar," Civin said.

More recently, however, Civin credited Zerhouni with helping remove obstacles to stem cell research. For starters, the NIH has published an online directory that clarifies which cell lines are available - nine in all - and tells scientists whom to contact. The agency also has set up a lab where scientists will study each cell line and tell researchers how they differ.

"He's responded in a very productive way," said Civin, who has obtained samples of two cell lines since the Senate hearing. "He listened, took notes, and responded to every single point we've made."

Even now, however, Zerhouni refuses to say if he shares the president's moral reservations about using human embryos.

"Well, you're applying a litmus test to me now. When you are a federal official in this position, I think you should really keep your own opinions outside the debate. The director of NIH should be factual, not factional."

As director of a sprawling bureaucracy with 27 components and a $23 billion budget, Zerhouni has plenty on his plate.

Among his chief goals is to break down some of the walls that separate scientists who labor in academic departments, inside and outside NIH. Categories such as cardiology, cancer or even chemistry made sense years ago - before the genetics revolution made clear that similar mechanisms are behind diseases that once seemed vastly different.

"The research you need to do on the cardiovascular system is not that far from research on the neurological system," he said. "For example, the concept of cell signaling" - how cells communicate with each other - "is the same whether it's in the heart or muscle."

Zerhouni said he is troubled by the tendency of unproven therapies to take hold without evidence that they work. Recently, clinical trials showed that hormone replacement therapy does not help and can even harm women who use it for extended periods of time, and that arthroscopic surgery does not relieve the pain of an arthritic knee. Yet, doctors promoted the unproven treatments for years before clinical studies finally shot them down.

For 30 years, he said, doctors assumed that hormone therapy was helpful. "People feel good, look wonderful - it's helpful, saves the heart," he said, summing up old dogma. "Then we say, 'Where's the evidence?' and we put together a trial, and $700 million dollars later, you find out that was totally wrong."

Zerhouni said he wants to find a way to organize the nation's health care system so evidence can be gathered as doctors treat patients, who often move around and switch health plans. In a better world, he suggested, women taking hormones could have enrolled in trials through their private doctors, agreeing to periodic tests that would have answered key questions about benefits and risks.

Similarly, men with family histories of prostate cancer could register for studies that would track whether they develop malignancies over time, and would decide the merits of screening and treatments.

Sweden, he said, has done a good job at this, organizing women into huge clinical trials that have produced some of the best data on mammograms.

But he said the task there is easier because Sweden has only 8 million people and a national health care system, which makes data collection easier. His job will be to see how this can be done in a nation of 280 million with a fragmented insurance system.

"That's the vision I have," he said. "It will take years."