(Time Inc.) Ask Herbert Benson to talk about his latest experiment and the distinguished Harvard researcher clams up. "Don't even ask," he says. "The more we say in public about the study we're doing, the more risk we run of compromising the results."

Neither Benson's silent treatment, however, nor the fact that the study is still in the planning stages has prevented word of his investigation from sparking a heated debate in the journal Science.

Small wonder. Benson, who directs the Mind/Body Medical Institute in Boston, has decided to rush in where most scientists fear to tread. In an experiment rumored to include more than 500 heart patients, he hopes to prove what has long been considered a matter of faith alone: the healing power of prayer.

Across the country, in fact, a growing number of scientists are putting prayer to the test. At Temple University in Philadelphia, researchers plan to examine whether the prayers of volunteers can speed the development of premature babies. At California Pacific Medical Center in San Francisco, psychiatrist Elisabeth Targ is studying whether prayer helps AIDS patients. "There are many of us who believe that prayer has enormous power to heal," says psychiatrist David Larson, who heads the National Institute for Healthcare Research, a Maryland-based organization devoted to the study of religion and health. "Now, as scientists, we're looking for proof."

Many Americans, it seems, don't need convincing. According to one recent study, 87 percent of adults believe their prayers are sometimes answered. In a survey conducted by the journal Nature, two out of five scientists said they believe in a God who answers prayers.

Is science itself beginning to show that they're right? The answer, it turns out, is not so simple. Prayer can heal, the latest findings suggest. But not the way most people suppose.

Scientists have been scrutinizing prayer for more than a century, but the first study to make skeptics sit up and take note was published in 1988 by cardiologist Randolph Byrd of the University of California at San Francisco. Byrd had divided 393 heart attack patients randomly into two groups. One group got standard medical care only. The others also received the daily prayers of volunteers who described themselves as born-again Christians. Neither the patients, nor the hospital staff, nor Byrd himself knew who was prayed for and who wasn't.

Reviewing hospital records, Byrd found that patients singled out for prayer required fewer antibiotics, suffered less congestive heart failure, and were less likely to develop pneumonia. His conclusion fairly leaped off the pages of the medical journal in which it was published. "These data," Bird wrote, "suggest that intercessory prayer to a Judeo-Christian God has a beneficial effect in patients admitted to a coronary care unit."

Since Byrd's study came out, several researchers have ventured into the same territory, seeking to show that so-called remote prayers by one person can help heal another. So far the evidence is iffy.

Last year, for instance, when University of New Mexico psychiatrist Scott Walker tested whether remote prayer could speed the recovery of substance abusers, the results turned out to be a bust. For one thing, the patients who were prayed for by volunteers were no more likely to kick the habit than those who weren't. Walker also asked patients if any friends or relatives might be supplicating on their behalf. Those who said loved ones were praying for them were actually less likely to give up drinking.

Perhaps the prayed-for patients in Benson's proposed study, which is modeled on Byrd's, will fare better. Still, critics ask, what would even the most dramatic results really mean? "Frankly, it's hard to see how studies like this can tell us anything," says Dan Blazer, a professor of psychiatry and the dean of medical education at Duke University.

The problem, he says, is that prayer can't be tracked or measured. No matter how carefully controlled the experiment, there's no way to know which subjects are being prayed for and which aren't. True, volunteers may be recruited to intercede for only one group. But at least a few family members or friends are probably praying for the others. (Herbert Benson won't talk about his study for fear that people who hear about it might send unsolicited prayers toward his patients, potentially skewing the results.) One could argue that a legitimate scientific study has to measure the "dose" of prayer each subject receives--impossible without some kind of prayer meter.

As it is, critics say, you can interpret any kind of result any which way. "If a study finds no benefits, you can say there weren't enough prayers, or they weren't done right, or by the right people," says Hector Avalos, a former faith healer who is now an assistant professor of religious studies at Iowa State University.

Even negative results can become proof of prayer's power. Consider the study Walker did. You'd think the fact that patients fared worse when they knew family and friends were praying for them would argue against the notion that prayer works. But Walker offers a different interpretation: "Suppose I'm an alcoholic and you're my brother. I've called you up drunk in the night, you've had to bail me out, I've caused scene after scene. Chances are when you sit down to pray for me you're feeling a lot of anger and frustration. What we may be seeing here is the possibility that certain kinds of prayer have the power to do harm."

If Walker is right--and let's hope he isn't--trusting in someone else's prayers may not be such a great idea, especially if you're not on the best of terms with Aunt Martha or Uncle Ed. But that doesn't mean prayer cant help anyone. The surest way to tap its potency, studies suggest, may be by practicing religion yourself.

Indeed, there's growing evidence of the power of belief. In 1990 researchers at Northwestern University Medical School found that among elderly women who'd had hip surgery, those with a strong religious faith got back on their feet faster than did nonbelievers.

Other studies have shown that believers live longer: At Dartmouth-Hitchcock Medical Center in New Hampshire, heart patients who said they drew comfort from their religious beliefs were three times more likely to survive bypass surgery than those of little faith. In one of the latest and largest studies of its kind, epidemiologist William Strawbridge and his colleagues at the California Public Health Foundation tracked more than 5,000 people for 28 years. Death rates of men and women who said they attended religious services regularly were only 65 percent of those for people who rarely entered a church or synagogue.

Well, sure, skeptics say. Churchgoers are less likely to drink, smoke, or carouse. But when researchers figured in such obvious lifestyle differences, worshipers still lived longer and healthier lives.

Why? One explanation may be that a religious community provides strong social support. And the evidence suggests a network of friends and family is crucial to health. According to Strawbridge, social connections explained some of the advantages worshipers enjoyed, but not all. People who went to church were still healthier than those who went to play bridge.

In the Dartmouth study, heart patients with extensive social networks fared as well as those who found solace in their faith. But those with deep religious convictions and a strong community of friends and family were almost 14 times more likely to recover than those with few friends and little faith.

Psychiatrist Harold Koenig, director of Duke University's Center for the Study of Religion/Spirituality and Health, thinks belief in God may help people cope with stress and uncertainty. "People who believe in a compassionate, caring God take enormous comfort from that," he says. "They have a hopeful worldview that gives them a sense of control when things go wrong. And the more in control people feel, the healthier they are both physically and mentally."

Prayer itself may reduce stress much the way meditation does--by quieting the mind, slowing heart and respiration rates, reducing blood pressure, and lowering levels of stress hormones like cortisol. Dozens of studies have shown that such hormones can rob immune cells of their disease-fighting abilities.

As intriguing as these findings are, Dan Blazer, for one, doubts that scientists will ever find what Benson and Walker are searching for: irrefutable evidence that the prayers sent by one person can improve the health of another. It's not that Blazer, a former missionary who describes himself as a devout Christian, doubts prayer's power. "I pray for my patients every day," he says. "But that's not a matter of medical science for me. It's a matter of faith."