Who's Into Bestiality?

01The data

The survey. Aella's Big Kink Survey is a long, anonymous online questionnaire about sexual interests, with about 1.35 million cleaned responses, recruited by viral spread across the internet. Respondents rate dozens of interests on a six-point scale. The item this report is about reads: "I find bestiality, or sexual interaction with at least one non-human animals/amphibians/insects/birds/etc, to be: Not arousing / Slightly / Somewhat / Moderately / Very / Extremely arousing." The survey clarifies that it means real creatures — not being an animal yourself, and not fantasy creatures, which have their own sections.

The hardweight version. People who take a kink survey are kinkier than the public, so raw rates from this data run high. Hardweight is the version of the dataset built specifically to correct kink base rates. The method: find traits that go along with being kinkier, that the sample has too much of, and whose true population rates are known — then reweight until the sample matches those rates. It keeps respondents aged 14–50 from the US, Canada, the UK and Western Europe (527,931 people) and balances thirteen dimensions at once against census, CDC and polling benchmarks: sex, age, LGBTQ status, politics, ethnicity, BMI, childhood and adult sexual assault, religious upbringing, self-reported autism, ADHD and depression, and lifetime number of sexual partners. Because a few under-represented kinds of people carry a lot of weight, those 528k responses have the statistical power of about 42,000. Everything on this page uses hardweight unless it says otherwise.

TL;DRAbout 1 in 14 — and mostly "generic kinkiness," except the shame.

About one person in fourteen reports some arousal to bestiality — more men (9.1%) than women (5.3%), and far more bisexual men (23%). At first glance the interested look different in many ways: heavier porn use, more violent porn, earlier sexual starts, less straight, a bit more autism. But they are also much kinkier across the board, and most of the profile is simply what kinky people in general look like. Compared with equally kinky people, depression, anxiety and ADHD differences are zero, and the rest shrink.

What survives: shame about their own arousal (by far the largest), heavier and earlier porn use, a taste for violent porn (especially in women), lower heterosexuality, a small extra dose of autism — and slightly fewer sexual partners than their equally kinky peers. They also lean introverted, rate themselves as less attractive, and strongly prefer animated porn. Rates are lowest in the UK and highest in Germany and France, and nearly three times higher among trans women than cis men. Their other fantasies aren't random either: bestiality clusters with the taboo/creature kinks (incest, mythical creatures, transformation, horror) and runs against mainstream BDSM.

02How common is it?

7.3% report at least some arousal to bestiality (95% CI 7.0–7.5%), and 4.2% say "moderately" or more. Among men it's 9.1% (5.4% moderately+); among women 5.3% (3.0% moderately+). It isn't a long thin tail: among the interested the ratings are spread fairly evenly, peaking at "moderately," with about 1.2% of all men and 0.6% of all women choosing "extremely."

Arousal is lowest among the young and climbs through the twenties: about 6% of men under 23 versus 10–11% from the late twenties on, where it plateaus through age 50. Women follow the same shape at a lower level — 4% under 23, peaking around 6.4% at 27–34, then easing to about 5%. That could be aging, cohort, or growing honesty with oneself; the onset data below suggest some of it is real acquisition in adulthood.

Orientation matters a lot, mostly for men. Bisexual men are at 22.9% — two and a half times straight men (8.8%) — with gay men in between (13.2%). Among women, bisexual women stand out (8.9%) while straight and gay women are about the same (5.0% and 5.4%).

03Which animals?

Dogs and horses, overwhelmingly. Among the interested, 75% find dogs erotic and 67% horses; nothing else on the main list passes 12%. Men and women like dogs equally; horses skew male (70% vs 61%), as do sheep and goats. Lions and bears skew female (lions 16% vs 8%), and women are likelier to tick "more not listed here" (32% vs 23%).

Forced to pick a single favorite, dogs come first for both sexes (47% of men, 46% of women). Horses are next but more of a male pick (38% vs 29%), and women are twice as likely as men to name something off the main list (18% vs 9%).

The exotic list is led by octopi/squid (11% of all the interested, and 16% of interested women — tentacle erotica is a genre for a reason), then foxes (10%), dolphins/whales (7%) and reptiles (6%). Insects, parasites and slugs are rare but not zero, at 3–4% each.

04Which acts, and since when?

Women's fantasies are strongly receptive: 63% of interested women find being penetrated by their preferred animal moderately or more arousing and 50% say the same of receiving oral, against 19% for penetrating it. Men have no such pattern — all four acts land between 25% and 31%, with being penetrated (31%) slightly ahead of penetrating (25%). So the male version of this interest is not mainly about penetrating an animal.

For men, onset looks like classic fetish timing, peaking at 13–16. Women are more spread out: more of them date it to age 10 or younger (10% vs 6%), and more to adulthood — 34% of interested women say the interest began at 19 or later, versus 27% of men. A meaningful share of people acquire this interest as adults.

05Who are they? (before adjusting for kinkiness)

The profile of the interested against everyone else, hardweighted like everything else here:

In unadjusted comparisons (hollow circles in the chart in the next section) the biggest gaps are in sexuality itself: more violent porn (d≈0.55–0.67), more shame about their arousal (0.47–0.62), more frequent porn (0.5), earlier porn and masturbation (−0.24 to −0.32), and less heterosexual (−0.25 to −0.32). Mental-health differences are much smaller: autism and sociopathy around 0.1–0.15, depression and ADHD about 0.1 or below, anxiety nil. Childhood sexual assault is elevated among interested men (0.18) but barely among women (0.05).

Background matters less than you might guess. The rate sits between about 6% and 9% in every group we split by. It is highest among the very liberal (9.0%) but second-highest among the very conservative (7.9%), with moderates lowest (6.5%). It follows a shallow U-shape by upbringing — both sexually repressed (7.7%) and liberated (7.6%) childhoods run above the balanced middle (6.5%). It declines gently with childhood social class (8.1% low → 6.7% upper-middle and up), and is a little higher among those raised without religion (8.3% vs 7.1% raised Christian). The male–female gap holds in every group. Country and ethnicity are covered further down.

06The key question: what's unique, vs just "being kinky"?

Bestiality-interested people are much kinkier overall (1.60 vs 0.96 on the 0–5 composite). So is any of the above actually about bestiality, or is it the standard profile of a very kinky person? The solid dots below re-estimate every difference holding general kinkiness and age constant — comparing the interested to equally kinky people.

Depression, anxiety and ADHD: nothing — all at zero or slightly below it once kinkiness is held constant. Self-rated breadth of interests collapses too. These are correlates of being kinky at all, not of bestiality. What survives:

• Shame (d≈0.42 men, 0.56 women) — the most bestiality-specific feature, and barely dented by the adjustment. Even among equally kinky people, the interested are far more likely to agree "I am ashamed or embarrassed about at least some of what arouses me."
• Porn: heavier (d≈0.26–0.31) and earlier — they started watching porn and masturbating younger (−0.17 to −0.24).
• Violent porn — a residual taste that is twice as strong in women (0.34) as men (0.16).
• Lower heterosexuality (−0.25 men, −0.16 women) — the orientation effect is not just kinkiness.
• Autism (d≈0.08–0.09) — small, but it doesn't vanish. Self-reported sociopathy shows a similar small residual, clear only in men.
• Fewer partners (−0.14 men, −0.07 women) — the unadjusted difference flips sign: for their kink level, the interested have somewhat fewer sexual partners, consistent with a more fantasy-weighted, less enacted sexuality.
• Childhood sexual assault, men only (d≈0.09) — about half the unadjusted gap remains for men; for women there is no difference at all.

The shame gap holds at every level of kinkiness. Among people who aren't into bestiality, shame rises with kinkiness from 40% to 60%. Among those who are, it is 73–80% almost regardless — a barely-kinky person with this one interest is more ashamed than the kinkiest people without it. Daily porn use shows a steadier gap of 12–19 points at every level.

Does any of it scale with intensity? The chart below sorts people by how arousing they rated bestiality and asks what share of each group has some other trait.

• Violent porn scales steadily. Among men who rated bestiality "not arousing," 15% say a moderate amount or more of the porn they watch is violent. Among men who rated it "extremely arousing," 39% do. For women the same comparison is 23% versus 61%. (Asked only of people who watch porn.)
• Autism rises at the top. 4.5% of men with no bestiality interest report autism, against 12% of men at "extremely" (women: 4% versus 9%, with wide error bars).
• Shame is a step, not a slope. About half of people with no bestiality interest agree they're ashamed of some of what arouses them. With any interest at all that jumps to three-quarters, then stays flat however strong the interest.
• Depression barely moves.

07The kink signature: taboo, not BDSM

Which other kinks does bestiality travel with, beyond what general kinkiness predicts? Partial correlations, controlling each person's mean arousal across the other kinks:

The signature is a taboo/creature/transformation cluster. Incest is the standout (partial r=0.34 men, 0.30 women), then mythical creatures, transformation, creepiness and horror, genderswapped and abnormal bodies, and reproduction. The age-related items are part of the cluster too and are notably more male: arousal to pre-pubescent children correlates 0.22 in men against 0.07 in women. At the bottom, the most mainstream kinks — teasing, eagerness, spanking, light bondage, sadomasochism — correlate negatively once kinkiness is controlled. Bestiality is not "BDSM but more"; it sits on a different axis, one organized around violating category boundaries rather than around power and sensation.

The neighboring communities tell the same story. People into furries, animal transformation, or animal body parts are five to nine times likelier to report bestiality arousal (42–52%, against a 5–9% baseline); pet play roughly triples it. But the overlap runs mostly one way: only 18% of the bestiality-interested are into furries, 20% animal transformation, and 36% pet play. They are not simply furries by another name.

08Species profiles: dog people vs horse people

Grouping the interested by their single most erotic animal. Three groups are large enough to profile under hardweight:

• Dog people (the plurality; 36% female): the most intense — 40% rate bestiality very or extremely arousing — the most receptive, and the most ashamed (80%).
• Horse people (30% female): similar but milder — less intense (30%), a little less ashamed (75%).
• Exotic people (54% female): the only majority-female group, the least ashamed (69%), the least intense, and the likeliest to also be into furries (21%) and to fantasize about penetrating (28%).

All three groups are about equally kinky overall (1.56–1.64), so these are differences of flavor rather than degree.

09Country and ethnicity

• The UK is lowest. 7.4% of British men and 4.4% of British women, against 10.1% and 6.5% in the US. Comparing people of the same sex and age, the UK runs 2.1 points below the US.
• Germany and France are highest. About 12% of men in both; 7.6% of German women and 9.6% of French women. Germans run 1.6 points above Americans of the same sex and age, and France 2.3 points.
• The German gap grows when kinkiness is matched, to 2.6 points. Germans in the survey are slightly less kinky overall than Americans (1.02 vs 1.12 on the 0–5 composite), so they are high on this specifically.
• Italy and Sweden show no sex gap — women are as high as men — though both samples are small.

The UK and Germany results appear in all three versions of the data (raw, softweight and hardweight). France is high in all three but only clearly so in softweight.

• Native American respondents are highest: 14.3% of men and 11.6% of women, 4.7 points above White respondents of the same sex and age. They're also the kinkiest group overall (1.23 on the composite), which explains part of it; 3.3 points remain once kinkiness is matched.
• Black, Hispanic and non-East Asian respondents run lower than White respondents, by 1.8, 1.0 and 2.1 points. None of that is explained by kinkiness; the gaps are unchanged after matching.
• East Asian and White respondents are the same.

The lower rate among Black respondents shows up in all three versions of the data. The Native American result appears in raw and softweight; under hardweight the group is too small to say.

10Gender identity and body weight

Gender identity is one of the largest splits in the data. About a quarter of trans women (24.8%) and AMAB nonbinary people (23.7%) report some arousal, against 8.9% of cis men. Trans men and AFAB nonbinary people are both at 11.7%, against 5.1% of cis women.

Trans and nonbinary respondents are also much kinkier in general (1.3–1.6 on the composite, versus about 1.0 for cis people), and that accounts for roughly half the gap. Compared with cis people of the same birth sex, age and kinkiness, trans women and AMAB nonbinary people are still about 9 points higher, and trans men and AFAB nonbinary people 3–4 points higher.

Body weight has a clear gradient. Among men, 7.0% at normal weight report some arousal, rising to 11.5% among the obese; among women it goes from 4.0% to 6.4%. Underweight people are also slightly above normal-weight people, though with wide error bars. About 60% of the obese–normal gap survives matching on kinkiness.

11Personality

The survey measures each Big Five trait with two agree/disagree statements, so these are coarse. Two things stand out, and both get slightly stronger when kinkiness is held constant:

• Introversion (d≈−0.25 men, −0.19 women). Extroversion here is "I am the life of the party" minus "I am quiet around strangers."
• Lower self-rated attractiveness (d≈−0.26 men, −0.15 women), from "Compared to other people of your same gender and age range, you are…"

The rest is small. Interested men score higher on feeling powerless (d≈0.14; items like "If life is a game, then I'm losing"), with no such difference in women. Agreeableness and conscientiousness are slightly lower in both sexes (−0.08 to −0.10). Neuroticism shows nothing, which fits the absence of any depression or anxiety difference.

As plain rates: 11.7% of the most introverted men report some arousal against 7.6% of the most extroverted, and 13% of men who rate themselves moderately less attractive than their peers against 6% of those who rate themselves significantly more attractive. The attractiveness pattern is mostly a male one.

12Porn, fantasy and relationships

A screen of every other question in the survey turned up a cluster of differences in how the interested use porn, what their fantasies look like, and how they approach relationships.

• Animated porn is the biggest, and it's untouched by the kinkiness adjustment (d≈0.51 men, 0.46 women). This is likely practical: drawn and animated content is where this material exists.
• Less consensual fantasies (d≈−0.27 men, −0.43 women). Asked for "the optimal amount of consent in my preferred erotic scenarios," the interested sit further toward nonconsent. This is about fantasy; the question says so explicitly.
• Porn-induced fetishes, paying for porn, polyamory, and attraction to trans and nonbinary people are all higher, each by about d≈0.2–0.3 after adjustment.
• Women only: interested women are much likelier to be aroused by two men together (d≈0.36). The mirror-image effect in men is zero.
• Men only: interested men lean more toward written erotica (d≈0.21).
• Less acted upon. The interested are less likely to agree "I've acted upon or experimented with all the things that arouse me" (d≈−0.09 men, −0.15 women), alongside their having fewer partners.

Four of these as simple rates:

Among men who watch only live-action porn, 5% report some bestiality arousal; among those who watch mostly drawn porn it's 23%. Among people whose ideal fantasy is fully consensual it's 5% of men and 3% of women; at the nonconsent end it's 17–20% of men and 13–15% of women. It rises from 6% of very monogamous men to 21% of very polyamorous ones, and from 4% of men who say porn hasn't given them new fetishes to 17% of those who say it has given them totally different ones.

13Where they came from, and what doesn't matter

How people found the survey predicts a lot. Respondents arriving from Telegram (18%) and Tumblr (17%) have more than double the overall rate; FetLife (12.5%) and Discord (12%) are also high, then Twitter (9.5%) and Reddit (8.8%). FetLife's excess disappears entirely once kinkiness is matched — FetLife users are just kinkier in general. Tumblr's and Discord's don't: they stay 6 and 3 points above people of the same kinkiness who arrived some other way.

Things that turned out not to matter:

• Attachment style. Secure, anxious, avoidant and disorganized all land between 7.0% and 7.5%.
• Absent parents. 7.7% with an absent father against 7.1% with a present one; an absent mother makes no difference.
• Birth order and number of siblings. Nothing.
• Childhood spanking and childhood social class. Small unadjusted differences that vanish once kinkiness is matched.
• Neuroticism, depression, anxiety, ADHD. As above: nothing once kinkiness is matched.

Put together, the childhood-and-family variables mostly come up empty. What distinguishes this group is concentrated in their sexuality itself and in a few traits of temperament.

14Methodology & limitations

Source & measures

Aella's Big Kink Survey (GuidedTrack), about 1.35 million responses after cleaning. The bestiality item is gated: respondents first tick whether the "Bestiality/creatures" category arouses them at all; those who don't are scored "Not arousing." Questions about specific animals, acts and age of onset were asked only of those with some arousal, and those percentages are reported only within that group. The survey's arousal scale is coded non-linearly (0, 1, 2, 3, 5, 8) and is recoded here to ordinal 0–5.

"Most erotic animal" uses the respondent's explicit pick, or the single animal ticked when they ticked only one. "Exotic" means they chose "More not listed here."

Weighting

Hardweight exists to correct kink base rates: its margins are traits that correlate with kinkiness and have known population rates, so balancing them pulls the sample's inflated kink levels toward the population's. It covers ages 14–50 in the US, Canada, UK and Western Europe and rakes to thirteen margins: sex × age × LGBTQ status; sex × politics; LGBTQ × politics; age × politics; age × ethnicity; age × BMI; childhood sexual abuse by sex; adult sexual assault by sex; religious upbringing; self-reported autism, ADHD and depression; and lifetime partner count by sex and age. Weights are capped at 30. The full sample's effective n is about 42,000; the bestiality-interested group's is about 3,200. All confidence intervals are 95% and use the effective sample size of the specific group shown (Kish approximation), so they reflect the weighting's cost.

Models

General kinkiness is the mean arousal across 60 other kink items, from spanking to vore. Kink-adjusted differences are weighted least-squares coefficients on standardized outcomes, with kinkiness and age (each with a squared term) as covariates, fit separately by sex, with robust standard errors. The kink-signature chart uses weighted partial correlations controlling a leave-one-out version of the same composite. Because that composite is dominated by the common kinks, some negative correlation with them is built in; the ordering is more meaningful than the exact values.

Limitations

• Fantasy, not behavior. Worth repeating: no question asks about actual acts.
• Weighting corrects for the known correlates of kink, not the unknown ones. Hardweight removes the inflation that comes from thirteen measurable traits, but whatever else makes someone likelier to take a kink survey is still in there. On sexual items that can be checked against probability samples, this dataset still runs well above them even after weighting. 7.3% is the best estimate this data supports, and the true rate is plausibly lower.
• One section is softweight. Country and ethnicity use softweight (ages 14–36, effective n≈169,000), because hardweight is calibrated for the whole population rather than within each group. Everything else is hardweight.
• Many comparisons. Sections 9–13 come from screening a couple of hundred survey questions for differences. With that many looks, some small effects will be flukes; the large ones (animated porn, fantasy consent, gender identity, introversion) are far outside what chance would produce, but anything under about d=0.1 should be held loosely.
• Coarse personality measures. Each Big Five trait is two survey items, not a full inventory.
• Cross-sectional. "Autism correlates with bestiality arousal" says nothing about direction or mechanism.
• Kinkiness control is a composite. Controlling for mean arousal across 60 kinks is a strong but imperfect way to define "equally kinky"; residual confounding by unmeasured dimensions of sexuality is possible.
• Self-report, anonymous, with all the usual caveats — though anonymity is exactly what makes taboo items like this answerable at all.

Analysis pipeline: R (DuckDB, data.table, ggplot2); scripts and full results tables available on request.