The biggest study into the relationship between prayer and healing found that, on the whole, prayers offered by strangers had no effect on patients' recovery.
But heart surgery patients who knew people were praying for them ended up with a higher rate of post-operative complications, such as abnormal heart rhythms. Researchers suggested this was perhaps due to the anxiety of knowing others were praying for their recovery.
In the past decade, scientists have carried out at least ten studies into the effects of prayer, and some have shown that "intercessory prayer" - prayer by strangers - can accelerate healing in patients.
The new study, involving more than 1,800 patients over ten years and costing about 1.5 million, was intended to overcome flaws in the earlier investigations and end years of speculation in both the medical and religious communities.
In the study, three Christian groups prayed for patients, starting the night before surgery and continuing for two weeks. Volunteers prayed for "a successful surgery with a quick, healthy recovery and no complications" for specific patients - they were given only the first name and first initial of the surname.
The patients were broken into three groups. Two were prayed for; the third was not. Half the patients who received the prayers were told they were being prayed for; half were told that they might or might not receive prayers. The results showed prayer had no effect on complication-free recovery. But 59 per cent of the patients who knew they were being prayed for developed a complication, compared with 52 per cent of those who were told it was only a possibility.
Researchers found the most common complication among those patients was atrial fibrillation, an irregular heartbeat that can be affected by anxiety. They speculated that knowing about the prayers made many patients nervous. "It may have made them uncertain, wondering, am I so sick they had to call in their prayer team?" Dr Charles Bethea, a cardiologist and the study's co-author, said.
Religious leaders said the results, published in the American Heart Journal were unlikely to discourage the devout. More than a quarter of British people say they pray regularly, and about a third of churches and other places of worship offer prayer-healing sessions for their congregations.
The Rev Malcolm White, a prominent faith healer from the Methodist Central Hall in Westminster, London, said 15 years of offering prayers for the sick had convinced him of the power of prayer.
He said: "I don't know exactly how it works, but I know it brings a change in attitude and health. I do feel that faith makes someone open to the possibility of a higher power helping them and that can bring peace of mind, which we know enhances the immune system. I have no doubt that what we do helps."
The Rev Dean Marek, a hospital chaplain and one of the study's investigators, said the research should not be taken as an indictment of a belief in a supernatural power, nor should it rule out the possibility that offering support to the sick does not help them heal. "Our study was never intended to address the existence of God or the presence or absence of [God] in the universe," he said.
Hospitals were filled with "thousands of people" trying to help patients get better. "That's intercessory prayer," he said.