18–30
Highest lesbian/same-sex porn search, highest hentai, more visual porn overall. Multipartner and novelty themes are loud. Power-exchange is already common. BookTok spicy and audio apps sit in this band.
Mind
What they fantasize about, watch, read, and hear
Frequency is one question. Content is another. This page is the content: the scenes, the porn categories, the books and audio used during arousal, and whether race or pregnancy rewrites the list. (Pregnancy kink and orgasm mechanics also live on the Pregnancy page.)
30–40
Same skeleton, denser flesh: more partner-narrative, more 'taken time with,' more MILF/amateur/reality porn, slightly more BDSM endorsement with experience. Written erotica holds or rises. Visual porn share falls a little.
Pregnant
Frequency follows the trimester U-shape. Content shifts toward the known partner and the pregnant body. Media mix tilts to imagery and audio. Fetal sex does not rewrite the list.
Fantasy themes
Share of women who report the theme at least occasionally. Joyal 2015 (799 women, 55 themes) is the statistical backbone: almost nothing on that list is rare. Lehmiller is larger and dirtier, and opt-in, so his percentages are treated as ceilings. A fantasy is not a request.
Current or known partner, with feeling
High confidence18–30: 88% · 30–40: 92%
The modal female fantasy is not a stranger. Joyal 2015 and Leitenberg & Henning 1995: romantic emotion during sex is typical (>84%) for women. Intensity of partner-focused scenes rises into the 30s.
Receiving oral sex
High confidence18–30: 82% · 30–40: 85%
Among the most common elaborated scenes in Joyal 2015. Matches Pornhub's female over-index on 'pussy licking' (+99% vs men).
Threesome / more than one partner
Moderate18–30: 72% · 30–40: 68%
Lehmiller (Kinsey, n > 4,000): group sex was the single most common theme, ~89% had fantasized about a threesome. Convenience-sample inflated; Joyal still rates it common. Slightly more often reported in the 20s.
Being dominated, tied, or overpowered
Moderate18–30: 64% · 30–40: 70%
Joyal: being dominated / tied is common, not unusual. Bivona & Critelli 2009: 62% of female undergrads had a 'rape fantasy' — almost always with a desired person, the woman still framing control, median ~4 times/year. Not a wish for real assault. Endorsement often rises with sexual experience into the 30s.
Sex with a woman (among mixed-orientation samples)
Moderate18–30: 58% · 30–40: 48%
Pornhub: women 18–24 are +31% more likely to search 'lesbian'; those searches drop with age. Many heterosexual-identified women have the fantasy without a corresponding identity. Not the same as orientation.
Public, caught, exhibition, new place
Moderate18–30: 54% · 30–40: 52%
Risk-of-being-seen and novel settings are common in both Joyal and Lehmiller. Pregnancy usually suppresses the acrobatic versions of this, not the idea.
Slow, romantic, 'taken time with'
Moderate18–30: 70% · 30–40: 78%
The 30s bump in fantasy intensity (Easton 2010, Purifoy 1992) is partly this: denser narrative, more lingering, less purely genital. Written erotica over-indexes here.
Taboo / role-play (boss, stranger, affair)
Low18–30: 46% · 30–40: 50%
Lehmiller's 'taboo' cluster. Affair and stranger-who-is-safe are more common than illegal themes. Joyal: sex with a child or an animal is rare (<2.3%) for women and men — those are not 'common kinks.'
What they use while aroused
Men over-index on visual porn. Women over-index on the mind, on prose, and increasingly on audio. Channels add to more than 100% because people mix them. Pregnant column is a trimester-averaged scenario.
Mental imagery only
18–30 42% · 30–40 48% · pregnant 55%
The default female masturbation aid is still the mind. No screen required.
Written erotica / romance
18–30 28% · 30–40 32% · pregnant 30%
Romance is a ~$1B+ market, ~80% women. BookTok 'spicy' titles concentrate in 18–34. Literotica, AO3, Kindle erotica.
Visual porn
18–30 38% · 30–40 30% · pregnant 18%
Rising in the 20s cohort; still well below men's rates. Pornhub 2025: women are 38% of visits worldwide — a platform statistic, not a census.
Audio erotica
18–30 18% · 30–40 14% · pregnant 22%
Dipsea, Quinn, r/GoneWildAudio. Women are the design target. Pregnancy and shared beds make audio (headphones, no screen-glow) relatively more useful.
Porn categories women actually open
Indexed to the top female cell (lesbian, 18–30 = 100). Source: Pornhub Insights 2025 plus the 2016 age split on lesbian search. This is what logged-in-as-woman traffic does on one platform — not a census of desire. Women over-index versus men on reality, 'popular with women,' and receiving oral. Lesbian search is a 20s story; MILF/amateur is a 30s story.
Race & ethnicity
Race does not choose the kink. Culture, religion, and who is allowed to say it out loud do.
There is no high-quality study showing that ancestry predicts a woman's fantasy menu after you hold education, religion, acculturation, and willingness to disclose constant. The seven Lehmiller themes show up in every demographic he could measure. What does differ: whether she will tell a survey she masturbates, whether the porn industry has a racial tag aimed at men, and whether she is fetishized rather than desired. Those are social facts. They are not a biology of kink.
| Group | Masturbation reporting | Porn | Kinks | Evidence |
|---|---|---|---|---|
| White (non-Hispanic, U.S.) | Highest reported past-month rates in NSSHB-family surveys; the reference group most papers compare against. | Broad category mix; default in Western platform data. | Lehmiller's seven themes (multipartner, power, novelty, taboo, passion, same-sex flexibility, nonmonogamy) all appear. Not a unique menu. | Moderate Reference category in U.S. papers. Higher reported masturbation is partly lower religiosity and higher social permission to say it, not a special kink genome. |
| Black (non-Hispanic, U.S.) | Less likely than White women to report recent/frequent masturbation after some demographic controls (Herbenick 2023). Partnered-sex prevalence is not lower. | No good evidence of a distinct 'Black women's porn diet.' Industry 'ebony' tags are marketed to men. | Same seven themes. BDSM-community studies find discrimination against people of color at events (≈16×), not a different fetish list. Ariane Cruz: race-play is a contested niche some Black women use and many reject. | Moderate Religiosity, respectability politics, and under-reporting explain more than ancestry. Being fetishized by partners is an experience of racism, not a map of her own fantasies. |
| Hispanic / Latina (U.S.) | Mixed. Some samples sit between White and Black reporting. Acculturation splits the group more than the label does. | Spanish-language and 'latina' tags are huge on platforms — again, mostly male search. Latina women in U.S. samples use both English and Spanish erotica. | No distinct biological kink profile. Catholic upbringing vs. second-generation urban acculturation is the actual fork. | Low Language, religion, and immigrant generation. The ethnic label is a bad proxy. |
| East / South Asian (U.S. & East Asia) | Lower reported rates in several U.S. and East-Asian surveys. Japanese and Korean published frequencies sit 30–50% below Western ones. | Hentai / Japanese is a top category for women 18–24 on Pornhub globally — that is a genre, not 'what Asian women watch because they are Asian.' Many of those viewers are not Asian. | More conservative attitude scores (Okazaki 2002 review) shrink with acculturation and with leaving the parental home. Fantasy themes, when asked privately, still include partner, oral, power, and multipartner. | Moderate Confucian/religious sexual scripts and under-reporting, not a race-linked fetish. Culture moves; chromosomes in this context do not. |
Industry tags like 'ebony,' 'asian,' and 'latina' describe what men search for, not what women of those groups fantasize about. Confusing the two is how fetishization gets mistaken for data.