The God question

What Prayer Has Cured

Vaccines ended or nearly ended diseases that killed American children by the thousands. The best-built test of prayer for the sick found no benefit beyond the medicine.

The point in one sentence

James promises that “the prayer of faith shall save the sick.” The best controlled test of that promise followed 1,802 heart bypass patients. Intercessory prayer, which means praying for someone else, made no difference to their complications. Over the same century, a worldwide vaccination and surveillance campaign took smallpox off the earth, and in the United States vaccines cut disease after disease by 95.7 to 100 percent.

The world this promise came from

If you grew up in the Church, you know what it is to be sick and to be prayed for. Two elders come to the bedroom. They anoint with consecrated oil, lay their hands on the person’s head, and give a blessing. A name may go on the temple prayer roll. Nobody who has watched a family do this thinks it is empty. It comes from love, and it is one of the most tender things the Church does.

It rests on two passages you may have heard in a lesson. The first is in the Bible, and the second is in the Doctrine and Covenants, the book of revelations to Joseph Smith.

What the scriptures and the Church say

The exhibits

Here is what the best studies found, in the researchers’ own words.

In the STEP trial, patients were randomly placed in one of three groups, and prayer was provided for two of the groups for 14 days, beginning before surgery. It was prayer by other people, not a blessing given at the sick person’s own request.

The column that is full

Now the other side of the ledger. These are numbers from the CDC’s own count, in a 1999 CDC report, Achievements in Public Health, 1900–1999: Impact of Vaccines Universally Recommended for Children, United States, 1990–1998. It compares each disease’s yearly cases in the years just before its vaccine (a different span for each disease) with the 1998 count, in the United States. The bars below are drawn on true scale: the top bar is 100 percent, a disease that is completely gone.

The baselines, by disease: smallpox, the yearly average for 1900–1904. Diphtheria, 1920–1922, three years before its vaccine. Whooping cough, 1922–1925, four years before. Tetanus, 1922–1926, estimated from deaths. Polio, 1951–1954, four years before licensure. Measles, 1958–1962, five years before. Mumps, 1968. Rubella, 1966–1968, three years before. Hib, an estimate from surveillance studies before the 1985 vaccine.

Smallpox is the one that is gone from the whole earth, declared eradicated by the WHO in 1980 after years of worldwide immunization and surveillance. The WHO says “the last known natural case was in Somalia in 1977,” and that it was “fatal in up to 30% of cases.” The CDC puts it plainly: about 3 out of every 10 people who got it died.

The WHO says that cases of wild polio have “decreased by over 99% since 1988, from an estimated 350 000 cases in more than 125 endemic countries, to two endemic countries” (WHO fact sheet, April 2, 2025). Add it all up across 50 years, and a 2024 study in The Lancet estimated that vaccination “has averted 154 million deaths,” 101 million of them infants. That is a model, not a count, and the authors present it as an estimate.

Measles shows what these numbers meant for families. Before the vaccine, the CDC says, 3 to 4 million Americans were infected every year, 48,000 were hospitalized, and 400 to 500 died. The Health Claims page has more on this: Health Claims and the Same Test.

What medicine counted, and what the prayer tests counted

Here is how the STEP trial worked.

What is true on the other side

The strongest believing answer is not “prayer is magic.” It is what Elder Oaks teaches, and what D&C 42 itself says. It is worth hearing at its full strength.

• The scripture already pairs prayer with care. D&C 42 says the sick, even those who “have not faith to be healed, but believe,” are to be “nourished with all tenderness, with herbs and mild food,” and the elders are called besides. That is a revelation that asks for both the nurse and the blessing.
• God can work through the doctor. Brigham Young taught it is consistent “to apply every remedy that comes within the range of my knowledge, and [then] to ask my Father in Heaven… to sanctify that application to the healing of my body.” Elder Oaks adds that “the use of medical science is not at odds” with prayers of faith.
• The Church sided with medicine. The First Presidency, August 12, 2021: “We urge individuals to be vaccinated. Available vaccines have proven to be both safe and effective.”
• Prayer is not a vending machine. Elder Oaks quotes a father whose daughter died of cancer: “Our family’s faith is in Jesus Christ and is not dependent on outcomes.” Elder Oaks adds: “We do all that we can for the healing of a loved one, and then we trust in the Lord for the outcome.”
• Blessings comfort people. A hand on a head and a calm voice can ease fear, and many families say that is what a blessing is for. The trials counted deaths and complications. They did not measure comfort.

All of that can be true, and it leaves two things standing.

First, prayer adds nothing that can be measured beyond the medicine. In the STEP trial every patient had surgery and the best hospital care. The only difference was the prayer. The difference in complications was 52 percent against 51 percent, which the authors counted as no effect.

Second, the promise in the text is a different promise. James says the prayer of faith “shall save the sick.” D&C 42:48 says that one with faith “shall be healed.” If prayer is for comfort and trust, then that is not what these verses say.

When prayer is used instead of medicine

Praying and getting treatment together takes nothing from the patient. The harm comes when prayer takes the place of treatment. The best-known study of this is a 1998 report in Pediatrics by Seth Asser and Rita Swan. They looked at 172 children who died between 1975 and 1995 in families from religious groups that rely on faith healing instead of medicine. Their finding:

140 of 172 children died of conditions that, by the authors’ estimate, children survive more than 90 percent of the time with medical care. Source: Asser and Swan, Pediatrics, 1998.

The study does not say these parents were careless, and nothing in it says they were reading James 5. What it shows is what happens when prayer takes the place of the doctor: the children died of conditions medicine usually treats.

The common rebuttals

“God works through doctors. The vaccine is the answer to the prayer.”

This is what the Church itself teaches, and it may be how it feels from the inside. But look at what it admits. If the answer always arrives as medicine, and the prayer adds nothing the trials can find, then the work you can measure is the work of the researchers and the nurses. In the STEP trial, every patient got the medicine. The only difference was the prayer, and it made no difference to complications. And James does not say “God will work through the doctor.” He says the prayer of faith “shall save the sick.”

“Vaccines prevent. Prayer heals. They are different.”

That is fair. Nobody prays to prevent smallpox. That is why the page does not rest on the vaccine column alone. STEP and the Cochrane review tested healing directly: patients who were already sick or facing heart surgery, prayed for or not. The vaccine column shows what happens when a remedy is tested and works. The trial column shows what happens when prayer gets the same kind of test.

“You cannot test God.”

Nobody tested God. They tested the claim that people praying for a patient changes how the patient does. That claim is in the text, and it is a claim about sick people, who can be counted. If the answer to every result is “that was the Lord’s will,” then the claim cannot fail, and a claim that cannot fail has nothing to say about whether to pray.

“I have seen a healing myself.”

You may have, and the relief was real. Most illnesses get better on their own, and people pray hardest when things are worst, so recoveries follow prayers whether or not the prayer does anything. That is why a trial compares a group that was prayed for with a group that was not. When that comparison was made on 1,802 patients, the prayed-for and not-prayed-for groups had the same rate of complications. Read more about how feelings and experiences are tested on How We Know Things, Believing Without Evidence, and The Case Against the God You Were Taught.

“A blessing is for comfort, not for cure.”

That is a fair thing to believe, and it is the version that survives. But it is not what the verse says, and a family that reads the verse as written is entitled to expect it. If comfort is the purpose, say so, and keep the doctor in the room.

What this means

You saw three things. Smallpox, a disease that killed up to 3 in 10 people, is gone, and the CDC’s own table shows polio, diphtheria and measles down by about 100 percent in the US. The best-built test of prayer for the sick, 1,802 patients, found no effect on complications, and the pooled review of 10 trials found no clear effect on death. And among 172 children who died in families that relied on faith healing instead of medicine, 140 died of conditions that, by the authors’ estimate, children survive more than 9 times out of 10 when they get medical care.

The promise says “shall save the sick.” Medicine ended these diseases. The prayer was tested alongside the medicine and added nothing anyone could find.

That is the finding, stated as exactly as it can be. The best-built test of prayer for the sick found no effect, and the Cochrane review of 10 trials found no clear effect on death. Two small early studies reported a benefit; the larger, better-built tests that followed did not find one. James 5:15 does not promise a small score difference. It promises that “the prayer of faith shall save the sick.” James wrote it as if it were certain. The question the evidence leaves is who made that promise, and why.