Why Do Cats Like Boxes - The Science of Sitting Small
Three separate lines of research explain the box, and none of them is the one you have read. Shelter trials, a temperature gap, and a square that is not there.
This page is not veterinary advice. Paw Multiverse is not a veterinary practice. We cannot examine, diagnose or treat your animal. There are no medicines here, no doses, no home remedies and no treatment steps, on purpose. If your pet is unwell, has swallowed something, or has changed in a way you cannot explain, ring your vet or your nearest emergency clinic and let them decide. Full disclaimer.
The question is not why cats like boxes, it is what a box is
A cardboard box gives a cat four walls, a floor, a roof of sorts, one exit, and a surface temperature that rises as the cat sits in it. Ask why cats like boxes and you get folklore. Ask what a box supplies and you get three separate lines of published research, two of which are unusually well designed for a question this daft sounding.
None of the three cancels the others out. A box can be a hiding place and a warm surface and an enclosing shape at the same time, and the honest answer is that all three effects are real and nobody has weighed them against each other in a single experiment. What follows is the evidence for each, with the numbers attached, and a clear line where the evidence stops.
Hiding is not sulking, it is a measurable coping behaviour
The strongest work on this comes from animal shelters, because shelters are where you can hand half the cats a box and withhold it from the other half without being cruel to anyone.
Claudia Vinke and colleagues ran that experiment and published it in 2014. Nineteen newly arrived cats were split into a group of ten with hiding boxes and a group of nine without, then scored over fourteen days using the Kessler and Turner Cat Stress Score, a standard behavioural instrument rather than a questionnaire. On days three and four the box group scored significantly lower, at p below 0.01. By day fourteen both groups had converged. The point is what happened in between: the box group reached at around day three the state the other group needed a fortnight to reach. The authors also noted much lower variance in the box group, meaning the boxes did not just help on average, they helped consistently.
A separate randomised controlled trial published in 2019 tested the same idea on 23 European shorthairs, eleven males and twelve females aged one to ten, split into twelve with boxes and eleven without, observed across twelve days of quarantine. Same direction, sharper numbers: the box group’s stress scores stabilised by day two, the control group not until day nine. Body weight moved the same way but did not reach significance, with the control group losing an average of 7.7 per cent of starting weight and the box group 6.3 per cent. Around 65 per cent of all cats in that trial lost five per cent or more. Adoption rates and length of stay were unaffected, which matters because it means the boxes helped the cats without costing the shelter anything.
Two independent teams, different countries, different sample sizes, same result. That is a stronger evidence base than most claims you will read about cat behaviour anywhere.
Why hiding matters more to a cat than it looks
A 2016 review of environmental aspects of domestic cat care sets out why. Hiding behaviour is correlated with an enhanced ACTH response and raised urinary cortisol, so it is not a mood, it is a physiological event that leaves a trace. More usefully, the review reports that thwarting a cat’s attempts to hide may contribute disproportionately to its overall stress. Removing the box is not neutral.
The same review cites a study in which cats given hide boxes were observed in or on them 77 per cent of the time, against 36 per cent of control cats attempting to hide without anywhere to do it. A cat with nowhere to go does not stop wanting to go. It just fails.
This is the practical takeaway that most box articles never reach. If your cat is in a box a great deal after a move, a new baby, building work or a new animal in the house, the box is doing a job. Taking it away because it looks untidy removes the coping option and leaves the stressor in place. Pet body language covers how to read whether the cat is settling or still braced.
The temperature explanation, which is better evidenced than it sounds
People roll their eyes at the warmth answer because it sounds like a shrug. The numbers do not support the eye rolling.
The thermoneutral zone for a domestic cat, the ambient range in which it does not have to spend energy warming or cooling itself, is 30 to 38 degrees Celsius, a figure the veterinary literature attributes to the National Research Council. Most homes sit closer to 22 degrees plus or minus two. That is a gap of at least eight degrees below the bottom of the cat’s comfortable band, permanently, in every room, all year.
One veterinary commentary translates this into a comparison worth holding onto: a home that suits a person can feel to a cat roughly the way a room at eight degrees would feel to us. Not dangerous. Just persistently, mildly wrong. A cardboard box is an insulator with a small air volume, and a cat sitting in one warms that air volume with its own body. The cat is not being eccentric, it is doing behavioural thermoregulation with the materials available.
This also explains the pattern most owners notice and dismiss: the box moves in and out of favour with the seasons and with which room the heating is on in. Nothing about the cat’s feelings changed. The physics did.
The illusion study, and what it does and does not prove
In 2021 Gabriella Smith and colleagues published a paper in Applied Animal Behaviour Science with the title If I fits I sits, which tested whether cats would sit inside a shape that was not really there.
The design used a Kanizsa square, an optical illusion in which four notched circles arranged at the corners of an invisible square make most human observers see a square that has no edges. Owners taped three kinds of stimulus to the floor at home: a real taped square, a Kanizsa square, and a scrambled control made of the same notched circles rotated so no illusory shape forms.
It was citizen science, and the attrition was brutal. Around 500 cats were recruited. Thirty finished the full six day protocol. Of those, nine selected a shape consistently, and among their selections the real square was chosen eight times, the Kanizsa square seven times, and the misshapen control once.
Smith’s reading is that cats are susceptible to the Kanizsa illusion in a broadly human way, and are drawn to the contours of a two dimensional shape rather than simply to novelty on the floor. Be careful how much weight you put on that. Nine cats making seventeen shape choices between them is a small pile of data, the authors present it as preliminary, and citizen science run in owners’ living rooms cannot control what the wider room looks like. What it does suggest, and what fits the box question, is that the edge is doing some of the work, not only the enclosure and not only the warmth.
Three explanations, one box
+-----------------+
| THE BOX |
+--------+--------+
|
+--------------------+--------------------+
| | |
v v v
+--------------+ +-----------------+ +----------------+
| ENCLOSURE | | TEMPERATURE | | CONTOUR |
| | | | | |
| one exit, | | cat comfort | | four edges |
| walls block | | band 30-38 C | | define an |
| approach | | home approx | | interior |
| | | 22 +/- 2 C | | |
+------+-------+ +--------+--------+ +-------+--------+
| | |
v v v
+--------------+ +-----------------+ +----------------+
| EVIDENCE | | EVIDENCE | | EVIDENCE |
| Vinke 2014, | | NRC thermo- | | Smith 2021, |
| n=19, and a | | neutral zone, | | n=30 finished, |
| 2019 RCT, | | cited in 2016 | | 9 chose |
| n=23. Stress | | welfare review | | consistently. |
| scores fell | | | | Preliminary. |
| days earlier | | | | |
+--------------+ +-----------------+ +----------------+
No study has pitted the three against each other.
Anyone ranking them is expressing a preference,
not reporting a result. Why the box beats the bed you bought
This is the part owners take personally. A heated, fleece lined, well reviewed cat bed loses to the box the bed arrived in, reliably, and it is not a comment on your taste.
Compare the two objects against what the research says a cat is looking for. The bed is usually open on all sides, which supplies no visual barrier and no single defensible entrance. It is often low and in the middle of a room rather than elevated or against a wall, and the 2016 review is clear that cats prefer to watch from height. It is large enough that the cat cannot warm the whole of it. And it has no edges at cat eye level to define an interior.
The box is worse in every way a human judges furniture and better in every way the literature says a cat judges space. If you want the bed used, the fixable variables are position, height, and whether the cat can see the room from inside it without being seen easily in return.
What we could not source, and therefore did not write
Several confident claims circulate on this subject with nothing behind them that we could find.
That boxes satisfy a hunting instinct because cats ambush from cover. Plausible, widely repeated, and we found no study testing it against the hiding or thermal explanations.
That the smell of cardboard glue attracts cats. Asserted often, sourced never, at least not anywhere we could reach.
That a cat in a box believes it is invisible. This is a joke that has hardened into an explanation. Nothing in the shelter work suggests cats misjudge whether they can be seen, and the Vinke results are about stress recovery, not about cats being fooled.
That box sitting indicates a happy cat. The shelter evidence points the other way if anything, because the cats using boxes hardest were the ones recovering from arrival. Heavy box use is not a warning sign on its own, but reading it as contentment is not supported.
When box behaviour is worth raising with your vet
This is not diagnosis and there is no treatment here. It is a short list of changes worth a phone call, because a vet can act on a change and you cannot.
Worth a call: a cat that has started hiding far more than it used to, a cat that hides and also stops doing things it used to do, a cat that hides in places it cannot easily get out of, a cat that goes to ground and stops eating or drinking normally, or a cat that hides and is also sleeping differently. Hiding paired with any other change is the pattern to report.
Describe what you have seen, when it started, and what else changed in the house. Before you call the vet sets out what to have to hand. We will not tell you what hiding indicates, because that is a consulting room judgement and we are not in the room.
Making the box do its job
If you are going to keep a cardboard box in your living room, and the evidence says you probably should, it is worth putting it somewhere it actually helps. Most of these adjustments cost nothing and take a minute.
Print it, or tick it here. It saves state until you reload.
Position
Fit and comfort
Household rules
Ring the vet if
Where this sits among the other things your cat does
Box sitting is one item in a set of behaviours that all trade on space, height and control of approach. Read together they tell you what your cat thinks of its territory.
If your cat also disappears for most of the day, the sleeping question covers what is normal. If it claims one particular chair and defends it, scratching is part of the same territorial picture. If it wants contact on its own terms only, licking, kneading and headbutting are the contact side of the same animal. Why cats hate water covers another case where an obvious explanation turns out to be shakier than the confident version suggests.
If you are setting a home up from scratch, the first thirty days with a kitten and pet proofing your home both matter more than any single piece of furniture. And if the hiding started because something in the house changed, a written care plan is a better tool than another shopping trip.
Where these facts come from
Every claim above was read on the source named beside it, on 4 August 2026. The 2014 shelter trial, the group sizes, the fourteen day window, the days three and four significance and the variance point come from the published paper by Vinke and colleagues, Applied Animal Behaviour Science, 2014. The 23 cat randomised controlled trial, the day two against day nine comparison and the weight figures come from the 2019 trial hosted on PubMed Central. The thermoneutral zone of 30 to 38 degrees, the 22 plus or minus 2 degrees household comparison, the ACTH and cortisol correlation, the point about thwarted hiding and the 77 per cent against 36 per cent figures come from a 2016 review of environmental aspects of domestic cat care. The eight degree comparison is from Veterinary Ireland Journal. The illusory contour study, the recruitment and completion numbers and the shape selection counts come from Smithsonian Magazine’s report of Smith and colleagues, Applied Animal Behaviour Science, 2021. Where a popular explanation had no source behind it, we have said so instead of repeating it.
