People really do recover, sometimes when doctors didn't expect it. If you've watched someone you love get better after prayer, that experience was real and so was your hope. This page doesn't call anyone a liar or a fool. It asks a simple question: when a person gets better after prayer, how can we tell whether prayer was the cause?
Bodies heal on their own far more often than we notice, and several well-known effects can make a recovery look like a miracle when nothing unusual happened. When researchers tested prayer by strangers for sick patients in large studies, they found no medical benefit. That doesn't test whether God exists, and it doesn't test whether prayer comforts people (it clearly does for many). It tests one narrow thing: does prayer change medical outcomes?
1. Why people get better, even without anything special
- Most illnesses end on their own. Colds, flu, most infections, back pain and many injuries get better with time. People usually reach for help (a doctor, a prayer) when they feel worst.
- Regression to the mean. When something fluctuates, like pain, a bad day is usually followed by a more ordinary one. If you act on your worst day, it will probably look like the action worked. Try it yourself below.
- Expectation (the placebo effect). Believing you're being helped can change how pain and nausea feel. A large review of 234 trials found that placebos don't have important effects on illnesses in general, but may affect things people report themselves, like pain and nausea.4 So hope can truly help people feel better. It doesn't shrink tumors or mend bones.
- Treatment is working too. Often a person is getting medicine and prayer together. The medicine may deserve the credit.
- Uncertain diagnoses. Sometimes a scary diagnosis turns out to be smaller or different than first thought.
- Rare real surprises. Doctors do occasionally record unexpected recoveries, called spontaneous remission. They're rare, and they happen to people of all beliefs and none. Unexplained means "we don't yet know why", not "we know it was a miracle".
- We remember the hits. We tell the story of the person who got better after prayer, and quietly forget the many who didn't. That's confirmation bias, which affects every one of us (see why we believe).
2. Try it: nothing was done, and the group still "improved"
Imagine 1,000 people whose pain bounces around every day between 1 (none) and 10 (the worst), just by chance. We pick only those who happen to be at 9 or 10 today (the worst days, when people ask for help) and look at how they feel tomorrow. Nobody gets any treatment at all.
Press the button.
This is a toy model (real illness is more complicated) but the effect is real: whenever there is random up-and-down, picking the worst moment makes anything done at that moment look good.
3. When prayer was put to the test
Scientists have tried to test prayer fairly. The best-known study (called STEP) followed 1,802 heart-bypass patients in six US hospitals. Patients were split into three groups, and the praying was done by church groups who got only patients' first names and last initials:1
| Group | Told | Actually prayed for? | Had complications |
|---|---|---|---|
| 1 (604 patients) | "You may or may not be prayed for." | Yes | 52% |
| 2 (597 patients) | "You may or may not be prayed for." | No | 51% |
| 3 (601 patients) | "You will be prayed for." | Yes | 59% |
Prayer made no difference to recovery (52% vs 51%). Patients who were told they'd be prayed for had slightly more complications, which the researchers thought might be due to the stress of expectation, though the study couldn't say why.
A broader review in the Cochrane Library (2009) looked at 10 trials with 7,646 people and concluded that prayer was "neither significantly beneficial nor harmful" for people who are sick, and that the evidence didn't support a recommendation for or against it.2
They test one specific thing: whether prayers by strangers, offered at a distance, change a medical result. They don't test whether God exists, whether personal prayer brings peace, or whether praying with someone at their bedside helps them cope, all of which many people find real. Some believers also object that God isn't something to test like a medicine. That's a fair point. The result is simply what it is: when the outcome is measured, no effect appears.
4. Lourdes: how a miracle gets "recognised"
Lourdes, in France, is the best-known pilgrimage site for healing. Since 1858, more than 7,000 healings have been reported there. The Church has formally recognised about 70 as miracles (the number has crept up by one or two in recent years).3 Nearly all of the thousands of reports don't pass its own review.
Its medical committee uses strict rules: the illness must be serious and recorded by medicine, show objective evidence (scans, tests), not be something that can be explained by treatment, and the recovery must be sudden, complete and lasting. Mental or nervous conditions don't qualify, so that the claim can't be explained by stress or suggestion.
The Church is careful, and for these cases doctors say "unexplained in the present state of our knowledge." Some of these recoveries had objective medical records.
"Unexplained today" isn't the same as "supernatural". Unusual recoveries happen everywhere, and this is roughly 70 cases from millions of pilgrims over more than 160 years. Nothing sets the cases apart that couldn't be rare natural events.
Both readings are consistent with the facts. Which is more likely depends on what else you know about how bodies, chance and evidence behave.
5. The missing-limb question
A common point skeptics make: recoveries after prayer seem to be the kind that sometimes happen naturally (cancers shrinking, fevers breaking, pain easing), but never a clearly documented regrown limb. Believers reply that God doesn't act on demand or according to our tests. Either way, it's a fair thing to notice, and it's why careful people ask what kind of recovery was seen.
6. How to look at a healing story: five questions
- Was there a real diagnosis? Tests, scans, doctor's records, or a feeling?
- What else was happening? Medicine, rest, diet, or a change in the illness?
- How likely was recovery anyway? Many illnesses get better by themselves.
- Did it last? And did anyone follow up months later?
- Has anyone else checked? A doctor who wasn't part of the story?
These questions aren't about doubting someone's love or sincerity. They're how doctors, and careful believers, decide what really helped.
Why do scientists compare a treated group with an untreated group?
- To be unkind to the untreated group
- To see what would have happened anyway, since many people improve on their own
- Because a bigger group always works better
7. A word about comfort
If prayer, church or faith helped you or your family through a hard time, nothing here takes that away. Comfort, community and meaning are real human needs. The aim of this page is only to separate "this helped us cope" (often true) from "this changed the medical outcome" (which, when tested, hasn't shown up). Doctors, nurses, family and prayer can all be part of a person's care at the same time.
8. Watch it explained
Large studies of prayer for sick patients found no effect on medical outcomes. Many recoveries have natural explanations. Some are still unexplained.
"I saw someone get better after prayer, so prayer healed them."
Many believers say, with reason, that the studies only measure what can be measured, that God can't be put to the test, and that their experience of healing, peace and answered prayer is part of life that no hospital trial captures. They also note that unexplained recoveries do exist. The reply: those are honest points, and "we can't explain it yet" is a fair place to stand. But a gap in our knowledge isn't evidence for any one explanation (see how science works), and when the same question is asked fairly, the result so far is no medical difference. Read next: why we believe and how do you know?
Sources
- Benson et al., "Study of the Therapeutic Effects of Intercessory Prayer (STEP) in cardiac bypass patients", American Heart Journal 151 (2006). Summary: Nature News report, 31 March 2006.
- Roberts, Ahmed & Davison, "Intercessory prayer for the alleviation of ill health", Cochrane Database of Systematic Reviews (2009). Cochrane summary.
- Sanctuary of Lourdes, "The miracles of Lourdes" (counts and criteria as stated by the Sanctuary; the number of recognised cases changes occasionally).
- Hróbjartsson & Gøtzsche, "Placebo interventions for all clinical conditions", Cochrane Database of Systematic Reviews (2010). Cochrane summary.