Skip to main content

Why Cats Love Earwax

, ,

Collation of anecdotes and speculation about why many cats like earwax, and human earwax especially. Is it the valeric acid?

2019-11-05–4y2024-08-21 in progress certainty: highly likely importance: 1 backlinks similar bibliography

While petting cats, I accidentally discovered about half of cats are fascinated by the smell & taste of earwax, particularly that of humans (both dry & wet), and this interest can last indefinitely. Dogs & humans, for comparison, are not. A number of anecdotes have reported this over the years, but no formal research appears to have been done on this.

What makes earwax attractive to cats? Pheromones? Some nutrient?

The best candidate to date is valeric acid, which is present in both human earwax & the cat attractant plant Valerian.

I accidentally discovered, by gently massaging cat ears while petting them, that cats can enjoy fingers inserted into their ears—perhaps because, like humans, earwax can build up to uncomfortable levels (and I once discovered undiagnosed ear mites in a kitten this way). This was how, after testing about 7 cats, I then discovered that cats are moderately interested in the smell/taste of their own earwax. (I’m not entirely sure if they exhibit a Flehmen response, as Kolb 199133ya claims. Most of my tests occurred before I learned what a Flehmen response was.) This has held true of most cats I have tested this on. Dogs, on the other hand, typically neither appreciate fingers in ears nor show much interest in smelling or licking any kind of earwax.

Curious, I tested other easily-obtained earwaxes on cats, and I’ve found that cat earwax is not even the most interesting earwax: cats are more interested in dog earwax, and human earwax most of all: the reaction can be quite strong—given the opportunity, a cat will lick and chew a hearing aid for a while, and I’ve had to take away hearing aids from my cats because the intensity of their licking made me worry about them damaging it. Aside from the cost of replacement, this is a safety concern: hearing aids, earphones, or disposable foam earplugs are small enough to be eaten & endanger cats’ health.

Searching, I’ve found a number of comments in publications and online also noting this phenomenon. There are not many contexts a cat owner would notice cats’ interest in earwax, and many of them are actively discouraged (a cat licking your ear hurts!), but when there is, it appears that earwax interest is often noted.

I first thought that it might be like sniffing butts, a way to learn about the health/status of another cat1, but that doesn’t explain why dog & human earwax is more interesting, and cats don’t seem to seek out earwax even after they know about it (except for one of my cats who’d sometimes try to lick my ears when I was in bed). I find that I sometimes have to touch their noses with a waxy finger before they abruptly become interested (suggesting that whatever the odor is, it doesn’t travel far), and the earwax smell also appears to be fairly temporary as fresh earwax is much more interesting than old earwax, suggesting that the odor also is temporary and volatilizes away quickly (so it’s not a long-lasting chemical inside the earwax which requires eating).

Chemistry

What is it about earwax that fascinates cats? Is there a particular chemical responsible, like a fat or salt or butyric acid?2 (While there is no shortage of confident assertions that the reason is the “incredibly high concentration of fatty acids and cholesterol” or salt in earwax and similar claims, exactly zero evidence is ever offered for these explanations.)

Earwax in humans comes in ‘dry’ & ‘wet’ forms, differing by race and I would describe cat/dog earwax as being like ‘wet’ human earwax; do cats like ‘dry’ earwax as well? I didn’t run into any explicitly Asian or ‘dry’ anecdotes, so in 2024 I ran an online survey targeting East Asian cat-owners, and of the ‘dry’ participants who had noted any cat/earwax exposure, 26% (14⁄52) reported their cat was interested in the dry earwax (which was similar to the reported ‘wet’ rate), which would appear to strongly exclude the possibility of 0%, and also confirmed individual-differences in earwax response rate (≤50% of cats). “Identification of volatile organic compounds in human cerumen”, Prokop-Prigge et al 2014 notes that the VOCs do differ quite a bit by race-earwax type (typically less in East Asian)—butyric acid is common to both, but inasmuch as I didn’t see any result from my simple experiments, that may be irrelevant.

Prokop-Prigge et al 2014’s Table 1 lists the 46 major VOCs they found along with racial differences; some of them can probably be rejected out of hand as candidates: Limonene (no difference?) would presumably result in cats having a notable attraction to citrus fruits (they don’t); 2-butoxyethanol or acetic acid or 2-ethylhexanol (no difference?) are used in many cleaning/painting/food products, likewise for the almond-tasting benzaldehyde (difference), and given the lack of cat interest in human urine/feces, p-Cresol can probably be ignored despite attracting many other animals like mosquitoes or horses or bees. More intriguing is propionic acid (no difference), which WP describes as “body odor”-like and related to cheese (reminding one of cats & licking sweat), apparently often produced by gut bacteria (ie. smells like “guts of prey”?) and a fatty acid—immediately sounding relevant to feline dietary needs & preferences!—which can be used in animal feed but is not present in high levels in household products that cats would be exposed to regularly. Other VOCs listed of interest are: caproic acid, MSM, BHT, enanthic acid, & caprylic acid.

Valerian

Most striking, however, is the entry of “pentanoic acid”, better known as valeric acid (no difference), as in, the second-most famous cat psychoactive, Valeriana officinalis! This would explain everything: the compound is not terribly common on its own, we know cats are attracted olfactorily to something in valerian (not necessarily the actinidine which is usually stated to be valerian extract’s active ingredient), cats are attracted to sweat too which contains isovaleric acid, valerian does not have any known effect on dogs and has minimal effect on humans (it’s used as a herbal remedy but in vastly larger quantities), valerian evaporates quickly, the Flehmen response is precisely how a cat responds to valerian, valerian response varies greatly between cats so that predicts earwax interest will vary (of the 2 cats I have tested with both valerian & earwax, the valerian responder was interested in earwax, and the non-responder was not), while the amount of pentanoic is inadequate to trigger any noticeable psychoactive effect and the valerian response tending to look like ‘passivity’ with the cat not seeking it out.

This explanation makes several predictions:

  1. purified valeric acid should be attractive

  2. dried or aged earwax should be low on valeric acid and low on attractiveness

  3. the ordering of valeric acid levels in earwax should be cat < dog < human

  4. East Asian earwax should be roughly as attractive (as Prokop-Prigge et al 2014 finds similar valeric acid levels in both dry & wet earwax) (which it turns out to be in the survey)

  5. there should be a high correlation between valerian-responding cats & earwax.3

Anecdotes

Searching Google Scholar/Google, I’ve found the following anecdotes:

  • Arny 199034ya, “A Craving For Wax”: letter to Nature asking if anyone knew anything after noting

    I dismissed this as a curiosity until I found that our second Siamese cat also liked it. In fact, the second cat leaps on the bed in the morning hoping to be offered some. I mentioned this odd behavior to 3 other people and have learned from them that their cats also liked the wax.

    I emailed Arny in November 2019, and he said he “got a few replies from people who had noticed the same behavior with their pet.” but otherwise nothing useful. (I have also told him about the valeric acid coincidence.)

  • Kolb 199133ya, “Chapter 25: Animal models for human PFC-related disorders. The Prefrontal Its Structure, Function and Cortex Pathology”, a paper which notes in passing that:

    The most common components of the response pattern include approaching, sniffing, and touching the urine source with the nose, flicking the tip of the tongue repeatedly against the anterior palate behind the upper incisors, withdrawing the head from the urine, and opening the mouth in a gape or ‘Flehmen response’, and licking the nose. This behavioral pattern apparently allows olfactory stimuli to reach the secondary olfactory system, which appears to be specialized to analyse odours that are species-relevant. Cats show this response to urine of other cats, and oddly enough to humans, but they do not show it to urine of rhesus monkeys, dogs, rats, or hamsters. They also do not show it to cat fecal matter or cat fur, although they do show it to cat earwax!

  • Rinaldo 200717ya, “Augmented Fish Reality”, an interview with an artist who mentions:

    At one time we had 13 cats and 7 dogs. I began being very interested in communications with cats and dogs and the subtle body languages that animals use to communicate. We had a cat named Que tu bu who loved to lick the earwax out of our ears, which was a strange scratchy affair, though clearly a cat showing affection and love toward a human. Later, as a teenager I became interested in Marine Biology…

    Rinaldo also mentioned this in a 2016 essay:

    After minutes of stroking, Catabu would suddenly pop up on his back paws and place his front paws on my shoulder. He would then begin to probe my inner ear with his scratchy tongue. His whiskers tickled as he dug further, licking my ear slowly and deliberately. This was somehow a pleasurable experience, though his tongue was sticky. Cat behaviorists, would speculate he was claiming me as litter-mate. I think we were exchanging love and affection. This was my first trans-species experience. Here was a cat, finding pleasure in the taste of my earwax while we provided mutual affection. This cat/human relationship eft a lasting legacy and deep-probing questions for me about animal-human communication, symbiosis and the contemporary notion of the computer interface.

  • Lynch 200717ya, “‘No Writer Nor Scholar Need Be Dull’: Recollections Of Paul J. Korshin”, in a memoir of English professor Paul Korshin, recollects:

    At the Osage house, Paul revealed himself as a doting cat lover. He and Debra had adopted brother and sister tabbies he’d named Oscar and Sherwin for his undergraduate mentor at City College. Oscar, the color of an orange creamsicle, would jump into Paul’s lap and purr during the seminars. Paul would cradle him, saying, “He likes to be made much of.” Both cats had free run of his exquisite suits, though he kept lint tape handy to pick up the fur. When Paul met my tuxedo cat Edgar, I mentioned that Edgar was partial to earwax. Intrigued, Paul said, “I don’t know if I have any”, but he put a finger to his ear and allowed Edgar to lick off the spoils. In time, Gaylord and Holly would join Paul and Debra’s cat family. We exchanged Christmas cards over the years “from our cat house to yours.”

On the general Internet, some cat owners have noted this behavior:

East Asian Survey

In August 2024, I ran an Internet survey on Prolific of n = 250 East Asian cat-owners asking about earwax type & if they had observed an earwax response from their cat. A majority had dry earwax, and most had not exposed their cat to earwax; of those who did, a majority observed no response; and of responders, 26% of dry earwax owners noted a response vs 43% of wet. This confirms that cats have individual differences in earwax response, and that dry earwax can trigger a weaker response—consistent with the valeric acid theory.

After reading all the available anecdotes, one of the biggest questions I had was whether the earwax effect is exclusive to ‘wet’ human earwax, which is the most common type in the West where all these anecdotes are coming from; no one specified being East Asian or having the other ‘dry’ earwax type. If there was no earwax effect for the dry type, that would be important to know for the chemical implications (and as evidence against the valeric acid theory).

I would have used Google Surveys to run a translated survey in Japanese, but Google shut it down. 3 years later in August 2024, it annoyed me I had still not run into anything on this topic, and I asked on Twitter about East Asian survey options and someone suggested Prolific; I was familiar with Prolific from its frequent use in psychology studies as an upgrade over Amazon Mechanical Turk (which appears to be suffering from increasing bot/fraud/quality issues) and Spencer Greenberg’s experiments/writings, but I had never gotten around to trying it myself. This was as good a reason as any.

I signed up and was surprised to see that Prolific allowed screening by not just ethnicity but pet ownership as well, including cats. So I could easily run a cheap ($50) survey of just n = 250 East Asian cat-owners, and then after the loss from people not having exposed cats to earwax or having the wrong wet earwax, I would at least be able to confirm or rule out an earwax effect of 0%. (Which would correspond to dry earwax lacking the critical chemical; if it contained the secret ingredient X, then a priori we would expect the effect to be weaker, because the dry earwax has fewer volatiles, and so people would be more likely to see little or no effect from measurement error.)

Survey Results

Given the screening, I worded it as a simple 2-question multiple-choice survey (CSV results):

  1. Is your earwax ‘wet’ (brown/sticky) or ‘dry’ (grey-tan/brittle)?

    • wet [n = 109]

    • dry [n = 144]

  2. Has your cat ever been exposed to your earwax; and was interested in it?

    • No; N/A [n = 153]

    • Yes; No [n = 65]

    • Yes; Yes [n = 35]

As expected, most results were not useful, but enough were to answer the question: dry earwax reported 14 responses vs 38 non-responses or 26%, and wet 21 vs 27 or 43%.4 Thus, we can strongly rule out 0%: dry earwax does trigger an earwax effect.

An interesting secondary implication here is that earwax effect is not universal: it seems many cats—perhaps even the majority—do not show any interest in human earwax. (This trend was visible in the anecdotes, but this sample size is much larger and confirms it.)

And if we take the overall average rate of 40%, we get a rate which looks like you would expect for the cat psychoactives like catnip or valerian. For example, a rate of 40% is consistent with the best estimate of valerian response rate, 50% (from the experimental data of Bol et al 2017), but biased downwards by measurement error (from issues like owners having only a little bit of earwax at a time).

The logical next question is then: is there a correlation between valerian and earwax?

Unfortunately, valerian root extract is used so rarely by cat owners that surveys are unlikely to dig up enough cat-owners who have already experimented with both earwax and valerian, and would start to run into methodological issues (eg. most Internet survey responses on that topic claiming to be that extremely unusual cat-owner might wind up being liars).

So this would likely require experimental investigation similar to Bol et al 2017 of a large group of cats in-person, where one can administer both valerian & earwax exposures.

  1. The Humane Society of the United States Complete Guide to Cat Care, Christensen et al 200420ya, notes in its ear cleaning section that with multiple cats who are good friends, it might be unnecessary, because it is a “allogrooming (mutual grooming) task that cats seem to love doing for one another.” Because of social bonding or to learn about each other—or because earwax is interesting & another cat is the easiest source? (I’ve never owned cats which were good enough friends, apparently, to see this for myself.)↩︎

  2. Both butyric acid and salt would also explain the interest in licking sweaty hands/armpits.

    I tried to crudely test out the possibility of butyrate on my earwax-loving cat. A 200618ya chemistry forum discussion outlined a number of strategies for obtaining butyric acid, of which the second simplest was adding lye and the simplest was just rancid butter.

    Since I had no lye lying around, I made rancid butter by putting a few tablespoons of butter into a ziplock bag inside a tupperware and storing it on top of my hot water heater for a week and a half; after a day it had melted and separated into a yellow liquid & a solid white mass, and after a week it smelled bad. I stored it in my refrigerator.

    I tested my cat 3× by paired dish comparisons (similar to previous tests of water or ofod): offering 2 saucers of a tablespoon of rancid butter and butter which I heated simultaneously in my microwave for 10s. All 3 times, if he started with the rancid butter, he would eventually lick and eat a little of the rancid butter, but slowly and without enthusiasm, and switch to the fresh butter which he would eat more of and much more enthusiastically, and vice versa, always leaving more of the rancid than fresh. The earwax-style fascination was absent for both kinds of butter. It’s possible that the rancid butter had other breakdown products which offset the appeal of the butyric acid, but I would have expected different behavior if he was simultaneously attracted & repelled.

    I further tested ‘caramelized butter’ (butter heated up to a point where it browns, which can be sweetened and used as delicious cake frosting); my cat liked it, but another cat I tried was almost indifferent.

    For a followup experiment, I bought some lye and mixed 2tsp lye with a quarter stick of melted butter (vs another quarter, both left to cool to room temperature), which turned it darker yellow and gave it an odd smell (albeit not nearly as strong as the ‘rotten’ butter smelt), which eventually hardened to a cracked white crust, and tried it on my cat; again, he showed no interest in the treated butter compared to the normal ‘fresh’ butter.↩︎

  3. But an imperfect one. It’s possible the response may be conditioned by prior valerian use, and the cat’s interest is partially “this earwax smells like that interesting valerian extract stuff”. We should also remember that cats are fickle and easily stressed, and measurements on them are unreliable: a cat may well be interested or responding, but just refusing to do so on this occasion—a kittlesome methodological problem that has sabotaged past cat behavioral research, including catnip research.↩︎

  4. Are they actually different rates? But given the small sample size, a two-sided proportion test is just p = 0.12, although since it is in the predicted direction, arguably the one-sided test of p = 0.06 is more relevant.↩︎

Similar Links

[Similar links by topic]