Does Counting Sheep Actually Work? Science Counted, and No

Somewhere around 1 a.m. on a Tuesday, lying in a Brooklyn apartment that shares a wall with a guy who practices bass, I caught myself doing it: one sheep, two sheep, three sheep. It's the folk remedy we all inherited, passed down like a family recipe nobody ever actually tested. So let's test it. Counting sheep has the three qualities of most famous sleep advice: it's old, it's free, and it's charming. Unfortunately there's a fourth quality, which is that when researchers finally put it in a lab, it lost. Not to some high-tech intervention. It lost to literally thinking about nothing in particular.

The study that retired the sheep

Researchers at Oxford took a group of people with insomnia and gave them different mental assignments at lights-out (Harvey & Payne, 2002, Behaviour Research and Therapy). One group was told to imagine a relaxing scene in rich sensory detail: a waterfall, a beach, whatever worked, built out with sights and sounds and textures. Another group got classic distraction of the counting-sheep variety. A control group got no instructions at all. The imagery group fell asleep meaningfully faster than their usual baseline, about twenty minutes faster. The counting-style distraction group did not beat the people who were given nothing. Read that again: the most famous sleep technique in the English language performed about as well as no technique.

Why sheep fail

The problem is workload. Counting is too boring to hold the floor. Your mind is a crowded room, and at night the loudest voices are the unpaid invoice, the weird thing you said in 2019, and tomorrow's meeting. A monotonous count occupies roughly four percent of your attention, which leaves the other ninety-six percent free to return to the worry playlist. The researchers' interpretation was essentially this: an effective pre-sleep task has to be engaging enough to crowd out the thoughts that keep you up, but calm enough not to become stimulation itself. Sheep are calm, sure. Engaging, no. It's the cognitive equivalent of hiring a very polite bouncer who lets everyone in.

Counting sheep occupies four percent of your attention. The worry playlist happily takes the rest.

What actually fills the channel

The winning move in that study, imagery, works because it uses the same mental machinery your worries need. You can't vividly render a slow waterfall and also rehearse Thursday's presentation; the brain's visual workspace is single-occupancy. That's also why techniques invented since then keep converging on the same principle. Cognitive shuffling has you picture random unconnected objects, one per letter: engaging, pointless, sleep-shaped. Body scans walk your attention through your physical self, toes to scalp. And watching slow, richly textured visuals does some of the rendering for you, which is the lazy person's version and I mean that as a compliment. A few minutes of a Feelsy slime texture or a drifting particle scene with the volume low gives your visual channel something dense and pleasant to chew on, which is exactly the job the sheep were failing to do. The common thread is sensory richness without plot. No stakes, no narrative, nothing to win.

How to run imagery distraction properly

If you want the do-it-yourself version from the research, here's the protocol, and it takes zero equipment. Pick a scene you find pleasant and mildly interesting, not exciting. Build it in layers: what you'd see, then what you'd hear, then temperature, then texture. The detail is the active ingredient, so go granular. Don't picture a beach; picture the exact pattern the foam leaves when a wave pulls back. When your mind wanders to real life, and it will, don't grade yourself, just go back to the scene and add another detail. If pure imagination feels like effort at the end of a long day, pair it with breath: one slow 4-7-8 cycle to downshift, then back to the waterfall. The goal isn't to force sleep. It's to make the pre-sleep channel so pleasantly occupied that sleep sneaks in the side door.

If you must count, count smarter

Some people are counters by temperament, and I respect it; numbers are tidy and the night is not. If that's you, the fix isn't to stop counting but to raise the difficulty until the task actually occupies the room. Count backward from three hundred by threes. The subtraction is just hard enough that your brain can't run the worry playlist in parallel, which is the entire point, and just boring enough that it won't wake you up trying. Another version yokes the count to your body: count only your exhales, one through ten, then start over, and when you lose the thread, which you will, begin again at one without commentary. Losing count isn't failure; noticing you lost count means you drifted, and drifting is the destination. The pattern across every version that works is the same one the Oxford data pointed at. The task needs enough cognitive load to fill the channel, a rhythm slow enough to be a lullaby, and zero stakes. Sheep, as traditionally deployed, fail the first test. Arithmetic in the dark, oddly enough, passes all three.

The verdict

Counting sheep isn't harmful, it's just weak medicine. If it works for you, genuinely, keep it; you've essentially built a personal mantra and there's no reason to fix what isn't broken. But if you've been counting past three hundred and wondering why you're still awake, the answer is that you were given a tool from an era when the competition for your attention at midnight was a candle. Your mind needs a bigger, softer object to hold. Give it a scene, a shuffle, or a screen's worth of slow-moving slime, and let the sheep retire. They've had a long career.

Feelsy is a wellness product, not medical advice. If stress, anxiety or sleep problems persist, please speak with a qualified professional.