Has prayer been tested in a clinical trial?

Yes, more than once. Researchers have run randomized trials in which patients were assigned by chance to be prayed for by strangers, without knowing which group they were in.

The largest was published in the American Heart Journal in 2006. It enrolled 1,802 patients having coronary artery bypass surgery at six US hospitals, and prayer was offered for fourteen days beginning the night before surgery. Among patients who were told they might or might not receive prayer, complications within 30 days occurred in 52 percent of those prayed for and 51 percent of those not prayed for. The authors concluded that intercessory prayer itself had no effect on complication-free recovery.

A second large trial, published in The Lancet in 2005, randomized 748 patients undergoing cardiac procedures and reported that masked prayer did not significantly improve clinical outcome.

A Cochrane systematic review pooled ten trials with 7,646 patients and described the findings as equivocal, concluding that the evidence does not support a recommendation either for or against the use of intercessory prayer.

None of these trials studied cancer, and none of them can say anything about what prayer means to a person. They tested one narrow question about measurable medical outcomes, and on that question the answer has been consistent.