Can you feel someone staring at you? The science is stranger than a simple yes or no

You’re sitting in a café, or on a train, or at your desk. You’re not looking around. And then — a pull, a pressure, something that has no name but is unmistakable — you turn. And someone is watching you.

It’s one of the strangest ordinary experiences there is. Not frightening, exactly. Just slightly uncanny. The feeling that something in the world reached into your awareness and tapped you on the shoulder, even though nobody touched you and you couldn’t have seen anything.

Almost everyone has felt this. Surveys across multiple countries put the number at somewhere between 70 and 97 percent of people who report experiencing it at least once. It has a formal name: scopaesthesia, from the Greek for seeing and feeling. It’s been documented across cultures and throughout history.

What it hasn’t been, even now, is explained to everyone’s satisfaction. Because the moment you try to test it properly, things get complicated — and the complication itself is worth paying attention to.

Why your brain cares about eyes

Start with what we know for certain, before moving to what we don’t.

Your brain has dedicated, finely tuned machinery for detecting gaze. Specific neurons — identified in neurological studies — fire in response to being looked at directly. The amygdala, which your brain uses to register threats, is involved. So is the superior temporal sulcus, a region that processes social information. Gaze detection is fast, automatic, and happens below the level of conscious decision — you don’t choose to notice that someone is looking at you, you just find yourself having noticed.

This makes evolutionary sense in the most direct way possible. A predator’s direct stare is a specific signal: it has spotted you, and the window before it acts is narrow. Prey animals that could detect that stare early had a better chance of surviving than those who couldn’t. Psychologist Colin Clifford of the University of Sydney has suggested that when people can’t tell exactly where someone is looking, they default to assuming those eyes are aimed at them — because in evolutionary terms, assuming you’re being watched and being wrong costs you almost nothing, while assuming you’re not being watched and being wrong could cost you everything.

So the bias toward feeling watched isn’t a quirk or a superstition. It’s a feature. The question is whether it extends beyond the range of ordinary sensory information — whether you can feel a stare you genuinely couldn’t have seen, heard, or inferred.

The researcher who spent decades testing it

Since the late 1980s, the biologist Rupert Sheldrake has been running experiments on exactly this question. His method is simple: one person sits with their back to a second person, who either stares at the back of their neck or looks away, in a randomized sequence. The seated person guesses whether they’re being stared at for each trial. No mirrors. No sounds. No feedback during the experiment in many versions.

Across tens of thousands of trials — including classroom experiments in schools in Germany and the United States, and a computerized setup at a science museum in Amsterdam where over 18,000 members of the public participated — Sheldrake’s results consistently come out above chance. Not dramatically, but measurably and significantly. His conclusion: something real is happening that science hasn’t yet explained.

Skeptical researchers pushed back hard. David Marks and John Colwell, writing in the Skeptical Inquirer, argued that the positive results were an artifact of the randomization sequences Sheldrake used — that the sequences weren’t truly random and that participants could unconsciously learn to detect patterns in them, producing above-chance scores without any actual stare-detection taking place.

Sheldrake’s counter: the positive results also showed up in trials using coin-toss randomization (which is genuinely structureless), and in trials where no feedback was given to participants at all — conditions under which the pattern-learning hypothesis couldn’t apply. The Amsterdam museum experiment used computerized randomization with no feedback, and still showed significant results.

The debate went on long enough to fill a dedicated issue of the Journal of Consciousness Studies in 2005, with fourteen scientists weighing in. It didn’t resolve.

What the skeptics get right, and what they don’t settle

The skeptical case is substantial. Sheldrake is not a neutral investigator — he has strong theoretical commitments to the idea that minds extend beyond brains, and some critics argue that experimenter effects (subtle, unintentional influences on how experiments are conducted and reported) could account for his results. The randomization critique, while not fatal to his entire body of data by his own account, raised real methodological questions.

What the skeptical case doesn’t fully settle is the experience itself. Nearly everyone has had it. The experience is so common and so consistent across cultures that dismissing it as pure superstition — pattern-matching on environmental cues, confirmation bias, the occasional coincidence remembered and the many misses forgotten — requires those mechanisms to be doing a lot of heavy lifting.

That doesn’t mean scopaesthesia is real in the way Sheldrake believes it is. It means the comfortable dismissal (“it’s just a superstition, nothing to see here”) and the overclaim (“this proves the mind extends beyond the brain”) are both stepping over actual uncertainty.

The honest position is narrower and stranger: we have a well-understood gaze-detection system that explains a lot of this, a body of experimental evidence that’s genuinely contested, and an experience so universal that it almost certainly isn’t nothing — even if what it is hasn’t been pinned down.

The question the feeling opens

Here’s what makes this more than a curiosity about turning around in cafés.

The mainstream neuroscience explanation works beautifully for gaze detection within normal sensory range — peripheral vision, subtle sounds, social inference running faster than conscious thought. All of that is real and well-documented. The interesting edge case is what happens when those channels are genuinely closed off. That’s where Sheldrake’s work sits, however contested.

And the reason that edge case matters is what it implies. If perception sometimes operates beyond the known boundaries of the senses — even slightly, even rarely — it would suggest that the model of mind as a closed loop inside the skull is missing something. Not paranormal. Not mystical. Just incomplete.

Sheldrake frames it that way deliberately: not as evidence of the supernatural, but as evidence that our picture of perception is narrower than the phenomenon itself. His critics think the experimental design is the problem, not the theory. The gap between those positions is the part that hasn’t closed.

The next time you feel that pull and turn around, you’re experiencing something that’s been catalogued across history, across cultures, across dozens of experiments — and that still doesn’t have a clean resolution. Your gaze-detection system is ancient, fast, and extraordinarily sensitive. Whether it reaches further than the senses can account for is the question scientists are still actually arguing about.

That’s not a reason to believe anything in particular. It’s a reason to stay curious about a feeling you’ve had since you were small, and that nobody has fully explained away.