Where’s the Evidence?

For thousands of years, humans have claimed that supernatural beings act in the world: answering prayers, healing the sick, working miracles. Those are claims about reality, and claims about reality can be checked.

This page isn’t an argument that gods can’t exist. It’s a narrower and more honest question: when we’ve gone looking for the fingerprints of supernatural intervention, carefully, with good instruments and fair tests, what have we actually found?

What Would We Expect to See?

Start with expectations, because that’s how any fair test works. If a being with unlimited power regularly intervened in human affairs, the evidence should be easy to find. We’d expect prayer to produce measurable results beyond placebo. We’d expect hospitals to document recoveries that biology can’t accommodate: regrown limbs, reversed brain death, healings that startle the physicians. We’d expect sacred revelations to contain knowledge their human writers couldn’t have had. None of these would be subtle. A God who parts seas shouldn’t be statistically undetectable.

Prayer, Tested

The most direct test is intercessory prayer, and it has been run at scale. The STEP trial, published in the American Heart Journal in 2006, followed 1,802 cardiac bypass patients: some prayed for by congregations, some not, with most patients unaware of their group. The prayed-for patients recovered no better than the others; the only measurable difference was that patients who knew they were being prayed for did slightly worse. Smaller studies claiming positive effects have repeatedly failed to survive larger, better-controlled follow-ups. The full story, including what’s happening in the brain during prayer, is on our page What’s Happening in a Praying Brain?

Miracles, Investigated

Credit where due: the Catholic Church takes miracle claims seriously enough to investigate them, with formal rules for evaluation and medical review. The shrine at Lourdes is the famous case study: since 1858, millions of sick pilgrims have visited, some 7,000 have reported cures to its medical bureau, and after review, the Church has recognized about 70 as miraculous. Seventy, from millions, over a century and a half.

And the recognized cases fit a telling profile. They involve conditions that occasionally remit on their own (cancers, paralysis, blindness with intact eyes), never conditions that don’t. No amputated limb has ever regrown at any shrine of any religion. Notice two more patterns: as medical imaging and diagnostics improved, recognitions became rarer, and every religion’s pilgrims report cures from their own gods at their own shrines. If miracles were divine signatures, we’d expect them to exceed what natural remission can do, to survive better instruments, and to point to one God rather than all of them. They don’t.

A Scoreboard With One Team on It

Zoom out and the historical pattern is hard to miss. Lightning was Zeus’s bolt; now it’s atmospheric electricity we can predict. Disease was divine punishment; now it’s germs we can cure. Eclipses were omens; now they’re forecast to the second, centuries out. Mental illness was possession; now medications outperform exorcists. In every case ever resolved, the explanation moved in one direction: from supernatural to natural, never once the reverse. Not one phenomenon in history has gone from “we thought this was natural” to “it turned out to be a miracle.” We trace that one-way pattern in full on God of the Gaps.

The natural explanations also do something the supernatural ones never did: they predict. Understanding orbits gives us eclipse tables; understanding pathogens gives us vaccines. “A god did it” has never generated a forecast, a cure, or a technology. Explanations that work earn trust. Explanations that only ever retreat earn doubt.

When Absence of Evidence Counts

You’ll sometimes hear that absence of evidence isn’t evidence of absence. That’s true when nobody has looked. It stops being true when the search has been long, competent, and motivated, and this search has been all three. Billions of believers praying daily. Churches investigating cures for centuries. Scientists, many of them religious and hoping for a positive result, running controlled trials. If the effect were there, someone should have found it by now, because everyone has been looking for it, everywhere, the whole time.

When an easy-to-find thing isn’t found by a massive, motivated, centuries-long search, that absence is evidence. Not proof of impossibility; evidence, the same kind we rely on when we conclude there’s no elephant in the room because we looked.

Where That Leaves Us

The conclusion this supports is modest, and it’s the site’s whole approach in miniature. It’s not “gods are impossible.” It’s that after the most thorough search in human history, the claims of supernatural intervention that can be tested have failed their tests, and the ones that can’t be tested can’t count as evidence either. A reasonable person, believing in proportion to evidence, is left with no good reason to believe intervention happens, and every reason to keep asking for the one thing that would change the picture instantly: a result that holds up. For who carries the burden of providing it, see Who Has to Prove It?

Want to Go Deeper?