So You Want to Know If Cats Can Actually Do Math
I spent too many years watching animal cognition research get sensationalized. The headline always says something dumb like "Cats Count!" and then the actual paper is about whether they can distinguish between two treats versus three. But there's a real answer here, and it's more interesting than most people think. The short version is that cats can process small quantities. We're not talking about long division. We're talking about subitizing — the ability to instantly recognize that there are three things without counting them one by one. Dogs can do this too. Rats can do it. It's a pretty basic survival skill, honestly. A 2019 study from Poland had researchers present cats with video clips showing different numbers of objects. One group saw a bowl go behind a screen and then one cat walk behind it — so two cats should be there. Another group saw one cat walk behind the screen, meaning only one should appear. The cats looked longer at the impossible outcome. That's the habituation paradigm, and it basically means they noticed something was wrong. They had an expectation about quantity that got violated.
Here's what nobody tells you though: the effect size is tiny. We're talking milliseconds of extra looking time. That doesn't mean the ability isn't real, but it does mean your cat isn't sitting there silently solving equations while you're on the Zoom call. I remember running a side project where I tried to train a rescue tabby I'd had for years to press levers for specific food amounts. She could reliably distinguish between one lever press giving one treat and two lever presses giving two treats. But as soon as I pushed it to four versus five, she just started guessing. That's the ceiling for most domestic cats — numbers up to about three or four, sometimes five if they've had specific training.
The Real Mechanism
Cats don't actually count in the human sense. What they're doing is more like an approximate number system, sometimes called ANS. It's the same system humans use before they learn formal math. You can tell "a handful of M&Ms" is bigger than "a couple" without really counting. Cats have something similar, just less precise than ours. What helps cats is that in the wild, quantity discrimination matters. You need to know if three mice is a fight worth taking on or a waste of energy. A single mouse might be worth going after. Litter sizes matter. It's practical, not academic. One counter-intuitive thing: cats actually do better with larger numerical differences. Distinguishing between two and four is easier for them than distinguishing between three and four. The ratio matters more than the absolute difference. This is called the Weber-Fechner relationship and it shows up across species. Bigger gaps are easier to perceive.
Get the Full Details
How to Test It Yourself
You don't need a lab. Here's the basic setup I used: Get two opaque containers. Put one treat in one and two in the other. Let your cat see you place them. Cover both containers with a towel or box so they can't see inside anymore. Point at each one and let the cat choose. If your cat picks the container with two treats more often than chance over about twenty trials, they're showing quantity discrimination. Keep sessions short. Five to ten minutes max. Cats lose interest fast and then you're just measuring their mood, not their math ability. Use high-value treats. Dry kibble won't cut it for most cats when you're trying to measure subtle cognitive preferences.
The edge case I hit most often: cats will learn the sound of treats shaking in the container instead of actually counting. I once thought my cat had improved at the task, but turns out she'd just learned that the louder rattle meant more food. I switched to using identical sealed bags inside opaque containers and the results became much cleaner. That's the kind of thing that goes unnoticed in amateur testing.
Where This Falls Apart
Don't read anything into your cat choosing the bigger pile. It doesn't mean they understand arithmetic. It means they have a rough sense of "more" versus "less," which is evolutionarily useful and cognitively basic. Calling that "math" is like calling breathing a sophisticated aerobic exercise. Also, individual variation is huge. Some cats will nail the two-versus-three distinction on their first try. Others won't bother engaging at all, which could mean they genuinely can't do it or just don't care about your plastic containers and treat promises. Motivation is the biggest confounding variable in any of these tests, and it's almost never controlled for outside of labs. If you're looking for a cat that can do actual math, you're looking at the wrong animal. But if you want to understand what basic numerical cognition looks like outside of humans and a few trained species, cats are perfectly adequate subjects. They're just not the stars of this particular show. Monkeys, crows, and some fish species actually outperform them on quantity tasks.
