Field of Science

It's official: praying for sick people doesn't help

Every few years, a group based at Hertford College at Oxford puts together a statistical analysis of all the studies conducted to date that have looked at whether praying for sick people helps them get better (or at least stay alive).

The latest has just been published, and it contains something pretty radically new in their conclusions: the evidence is now so clear cut that they think that no more studies should be done. The book is shut. Praying for sick people simply doesn't work.

Now, the odd thing is that there haven't actually been any new studies on this since their last report, back in 2007. So why the change of heart? There are a couple of reasons.

First off, this analysis is done under the auspices of the Cochrane Collaboration, which is an international group of experts devoted to pooling together the results of clinical trials to answer medical questions with unprecedented precision. The Cochrane Collaboration sets out the guidelines for the best ways to do this.

Last year, they upgraded their guidelines, a recommended a better statistical method (technical note: they used a random effects model this time, rather than fixed effects model). The previous analysis found a hint that praying for sick people might actually help them live longer. The improved analysis squashes that idea.

And the other new thing is some information on one of the studies they had previously included. This one looked not at death but the opposite: birth. The premise of the study was that people in the US, Canada and Australia prayed for couples undergoing in vitro fertilization. The result was, apparently, a doubling of the fertility rate in those couples who got prayed for - a fantastic increase.

I say 'apparently' because it turns out that the study was a fraud. Not only that, but the guy who ran the prayer groups was later jailed for an unrelated fraud. Strike one for the power of prayer.

Now, overturning the conclusions on technical grounds might make some people suspicious. Perhaps this is just a bunch of cynical scientists looking for an excuse to bury data they don't like.

But you'd be wrong. The lead author, Leanne Roberts, is not a scientist at all but in fact Chaplain of Hertford College. In previous editions of their analysis they were actually quite hopeful that they might see an effect.

All credit to them, they took a good hard look and the evidence and concluded that there was nothing there.

_______________________________________________________________________________________ Roberts L, Ahmed I, Hall S, & Davison A (2009). Intercessory prayer for the alleviation of ill health. Cochrane Database of Systematic Reviews (2) DOI: 10.1002/14651858.CD000368.pub3.

Creative Commons LicenseThis work by Tom Rees is licensed under a Creative Commons Attribution-Share Alike 2.0 UK: England & Wales License.


  1. I was a Christian for a certain portion of my life, and if you had asked me back then about studies showing that praying for sick people doesn't help them, my reply (at least in part) might have been roughly as follows.

    To study whether prayer helps sick people, one would have to intentionally arrange for some sick people to not be prayed for. God, disapproving of this, might well refuse to cooperate with the study, hence the null result.

    I wonder how you would have replied...

    1. I would reply that for god to kill one person to undermine the ideas of another is far from the image of compassion and justness religion loves to associate with him. So then if your example were the case it would prove empirically that god is either a heartless dick or the idea of him is fraudulent. Neither choice, however, would suit the religious picture theists attempt to paint.

    2. Then my reply would be that for god to purposefully kill the individuals whose families were praying for just to undermine those conducting the study would prove one of three things. Either: one, he is a fraud developed by man with no basis in reality. Two, he is not involved in the lives of man and has no preference on the outcomes of human interaction, in which case still would prove religion is still wrong in what it is they claim their God to be. Or three, he is a heartless d*ck that would kill innocent people just to undermine a scientific study which is merely attempting to grasp the world in which we live. I tend to lean toward one of the first two but after learning the story of Job it is truly hard to count the third possibility out.

  2. It's a fundamental problem. God can do whatever it wants, so any result can always be explained away. It's the attraction of religion, after all!

    In this particular case, whenever you pray for someone you are not praying for someone else. These studies test a particular idea about god - that god's healing power can be directed to some people but not others by human actions (prayer).

    To my mind the whole idea that God would withhold healing from you just because you don't have someone to pray for you is callous. But that's surely the premise behind praying for better health?

    If if is, then these studies show that premise is wrong.

  3. There was a similar Templeton study a while ago that also concluded no effect.

    Of course, here people also countered that God isn't willing to participate.

  4. You guys post the greatest stuff ! Thank you !
    I am a former Christian, Homeopath and Acupuncturist -- I have magically healed people in many ways.

  5. Do you have a link to that study??


  6. Jerome, Hi, Yes. The link's at the foot of the post:

  7. I'm sorry but that link doesn't seem to work?

  8. This study is flawed because people do not know how to pray correctly. Their prayers are not full of breath. The peoples who practice Huna as a way of life created miracles on a daily basis. Check the history of the missionaries in Hawaii and how they continued to fail to impress the people there. There is a whole other side that people do not know about. Case closed if you don't know what you are doing. This article and the study are fun for people to laugh who know better and case NOT closed at all.

  9. @Andrew. Is there an instruction manual for praying now? Also that's a mighty big assumption on your part about the people praying...

  10. It's simple, really:
    If your prayer is followed by healing, then you prayed correctly.
    If it is not, then you prayed incorrectly.
    So, all the study has shown is that most people pray incorrectly.
    But you needn't be one of them.
    For a small fee of $9.95 + shipping & handling, I can teach you how to pray correctly.

  11. We can expand on this as well. A while back I read a blog post in which the poster wondered how prayer can ever work. Let's assume (as most religions do) that a god is omnipotent: all knowing, all encompassing. When we encounter a situation which requires prayer, we pray and pray for God to intervene and throw his/her benevolence our way. We deserve it--we're good religious folks, really. Why shouldn't God create a miracle (or change his mind about this) just this once? The only thing is, our omnipotent God knows exactly what the outcome of this situation is. In fact, he's known for eons. He knows whether the outcome will go well for you, or not... We even pray that he "change his mind" when it looks as if he's given the ominous "NO" or "MAYBE" answer, when it is implicit that an omnipotent god cannot "change his mind". God will have known what he will do at any particular instant in time from day 1! Changing of one's mind requires a mortal limitation of not knowing at one point exactly what your decision may be at another point. Omnipotence forbids this.

    This is independent of any psychological implications which some claim may help say, a terminally ill patient live a month longer due to "the healing power of prayer". A god will have always known when and how said patient will live or die...and no amount of praying will have any effect on it.


  12. Using a simple prayer of faith I have seen cancer healed, MS, tumors, etc. All I know is that the prayers included the follow "magic" words. Here they are for free: Father, in Jesus's name, I ask you to heal (name of person) from head to toe and from fingertip to fingertip. I curse this infirmity to the root, and I ask You to do it to Your glory, that the witnesses here may know You are real. Amen. The prayer of faith is based on believing it will be done. Works for me. Seen the x-rays and doctor's reports.

  13. Father, in Jesus's name, I ask you to educate Anonymous from head to toe and from fingertip to fingertip. I curse this logical fallacy to the root, and I ask You to do it to Your implausibility, that the witnesses here may know You aren't real. Amen.

  14. That article is simply not true. Prayers are like balllons when you let them go some fly higher in the sky and are found , but others pop because it is just meant to be. Manpower can not keep someone alive ,if god says its time for them to go to heaven its time.all the prayer in the world could not change it .


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS