Pornography: what the research supports

Why this is difficult

Pornography research is unusually hard to interpret. Almost all of it is correlational; experimental designs are limited by ethics; self-report of both use and effects is unreliable; and the field is heavily politicised, with well-funded advocacy on multiple sides producing confident claims the underlying data do not support.

The honest summary is that a handful of findings are reasonably well supported, several widely repeated claims are not, and some important questions remain genuinely open.

Reasonably well supported

Use is common. Large representative surveys consistently find a majority of men and a substantial minority of women report use.

Unrealistic expectations. Content systematically misrepresents body appearance, duration, arousal timing, lubrication, cleanliness of anal sex, women's orgasm from penetration alone, and the absence of negotiation, contraception and awkwardness. Adolescents whose primary sex education is pornography report expectations that do not match partnered reality — a finding that argues for better sex education rather than for suppression.

Population-level sexual violence has not risen with availability. The prediction that mass availability of internet pornography would increase sexual violence has not been borne out; rape rates in most high-income countries declined or were flat across the period of steepest increase in availability. This is ecological evidence and does not settle individual-level questions, but it is inconsistent with the strongest causal claims.

Compulsive use is real for a minority. ICD-11 recognises compulsive sexual behaviour disorder as an impulse control disorder. It is characterised by functional impairment — missing obligations, escalating use, failed attempts to reduce — not by frequency.

Distress often tracks moral incongruence rather than use. Grubbs et al.'s work found that self-reported "porn addiction" was predicted more strongly by religiosity and moral disapproval of one's own use than by the amount of use. This matters clinically: for a substantial proportion of people who present with distress, the effective intervention addresses shame rather than behaviour.

Contested or unsupported

"Porn addiction" as a substance-like addiction. Neither the DSM-5 nor ICD-11 classifies it as an addiction. The neuroimaging literature offered in support is small, heterogeneous and inconsistently replicated, and reviews such as Prause et al.'s critiques have found the addiction-model claims outrun the evidence.

Porn-induced erectile dysfunction. Widely asserted, weakly evidenced. Some studies find associations; large representative samples such as Landripet and Štulhofer (2015) found little or no relationship between pornography use and erectile difficulty. Plausible mechanisms exist — conditioned arousal patterns, unrealistic comparison, anxiety — and confounding with anxiety and relationship factors is substantial. Where a man's erections are fine alone and not with a partner, the useful clinical work is on anxiety and stimulation patterns rather than on a disease label.

Escalation as an inevitable trajectory. The claim that use inevitably escalates to more extreme content is not supported as a general pattern; it occurs in some compulsive users and is not the typical course.

Uniform effects on relationships. Findings are mixed and moderated by whether use is concealed, whether it is solo or shared, and whether the partners disagree about it. Concealment and disagreement predict dissatisfaction far more reliably than the use itself.

Genuine concerns with good grounding

• Production ethics. Consent, coercion, age verification, payment and working conditions vary enormously across the industry. Trafficking and non-consensual content on major platforms have been documented, prompting significant policy changes at several large sites since 2020. Choosing ethically produced or performer-owned content is a meaningful action.
• Access by children. The average age of first exposure is early, frequently accidental, and there is no defensible argument that this is a good way for children to learn about sex.
• Non-consensual and image-based abuse content hosted on mainstream platforms.
• Recommendation algorithms shaping consumption in ways users do not choose and often do not notice.

Practical

• If use is interfering with work, relationships or wellbeing, that is the criterion that matters, and it is worth addressing regardless of what it is called.
• If distress comes from shame rather than impairment, treating the shame is more effective than escalating restriction, which the evidence on moral incongruence supports directly.
• Treat it as fiction. Performers are performing; the bodies, durations and responses are selected and edited.
• Concealment damages relationships more reliably than use. Honesty about it, and agreement on what is acceptable, is the more useful conversation.
• Talk to teenagers about it directly. Pretending it does not exist leaves pornography as the sole curriculum.

Sources

• Grubbs JB et al. "Perceived addiction to internet pornography and psychological distress." Psychology of Addictive Behaviors, 2015; and "Moral incongruence and compulsive sexual behavior." Psychological Bulletin, 2019.
• WHO. ICD-11, 6C72 Compulsive sexual behaviour disorder.
• Landripet I, Štulhofer A. "Is pornography use associated with sexual difficulties and dysfunctions among younger heterosexual men?" Journal of Sexual Medicine, 2015.
• Prause N et al. Critiques of the pornography addiction model, Archives of Sexual Behavior and related commentary.
• Ferguson CJ, Hartley RD. "Pornography and sexual aggression: can meta-analysis find a link?" Trauma, Violence, & Abuse, 2022.
• Wright PJ, Tokunaga RS, Kraus A. "A meta-analysis of pornography consumption and actual acts of sexual aggression." Journal of Communication, 2016.

Last reviewed: August 2026.