Research on sex drive differences finds a higher male average. People still vary widely within each group. The finding does not settle what causes the gap. Nor does it tell you what a given partner wants. A group average is useful, but it is not a verdict on your relationship.
What does the research show about sex drive differences?
It reports an average gap between the groups. The review pooled 211 studies and 621,463 people. Its Hedges g was .69, described as medium-to-large. That figure fell to .54, described as moderate, after the authors adjusted for biased responding (Frankenbach et al., 2022). Hedges g gives the gap between means in standard-deviation units. It is not a percentage, ratio, or personal probability. The means differ, but the groups overlap. The indicators were sexual thoughts and fantasies, desire, and masturbation, not coercion. The review does not establish hormone causes.
What can a hormone trial tell us?
A trial assigned 790 men to testosterone gel or placebo by chance. They were 65 or older, had low testosterone and symptoms, and used the gel for a year. Sexual desire and function improved in that selected treated group (Snyder et al., 2016). The trial was too small to settle treatment risks. This article does not recommend treatment. The trial did not test what causes the sex gap.
Are desire and aggression the same thing?
No. Desire and aggression are not one trait. They are measured as two distinct things.
A 2021 U.S. probability-sample observational study of men found a link. Impersonal sexuality and extreme pornography exposure were associated with sexual aggression (Wright et al., 2021). The associations held even when sex-drive indicators were included in the model. The measured drive indicators did not account for the whole link.
The study was observational. It does not prove one thing caused the other, and it does not show that desire is irrelevant. The model accounts for measured factors, based on its assumptions. Other factors or flaws in the measures could still affect the result.
Impersonal sexuality and aggression are not the same measure. Keep the study's measures separate from a judgment about any one person.
Why is the origin of the gap still unresolved?
A 2024 commentary proposed that setting could explain the gap (Conley & Yang, 2024). It pointed to the quality of sexual experience, narrow definitions of sex, sexual violence, and stigma. The authors argued that if these factors were fully accounted for, the gap could shrink or even disappear. That is a proposal, not a new experiment. This commentary did not demonstrate that the gap disappears.
In reply, the original authors agreed that social experience may help explain the gap (Frankenbach et al., 2024). A test finding no change in the gap across settings does not rule out social effects. They also rejected the claim that their work proved a biological cause. The observed pattern and its causes remain distinct questions.
The debate asks why a gap shows up. The critique does not erase the observed mean. The mean does not prove one cause. Both points can be true at once.
How do desire, arousal, and consent fit together?
Desire means wanting sex. Physical arousal is a body response. Consent is freely given, ongoing agreement. Behavior is what a person does. A person can feel physical arousal without wanting sex. A past yes is not a current yes. Anyone can withdraw consent, and silence is not consent (RAINN, 2026). In a hypothetical case, someone feels aroused but says no. The body response is real. The no still stands. These are safety principles, not a legal test.
How should a careful reader weigh a claim about sex drive?
Judge a claim against the design of the study. A confident sentence does not make the evidence stronger. The three designs below answer different questions.
| Study design | What it can address | What it does not settle by itself |
|---|---|---|
| Pooled group comparison | Whether group means differ, and by how much on the chosen measures | Why the means differ, or what a given person is like |
| Randomized treatment trial | Whether the assigned treatment changed outcomes in the studied group | Why a wider group gap exists, or effects in other groups |
| Observational study with adjustment | Whether a link remains after accounting for measured factors | Whether other factors explain the link, or which way cause runs |
Comparing means does not by itself test a cause. A trial can estimate the effect of a treatment assigned by chance. It compares groups, rather than just noting a change over time. A model can account for measured factors. It cannot rule out all factors that were not measured.
Now test three hypothetical claims.
First: "The average gap proves biology." Ask how the study tested that cause. Could social factors also help explain the result? A group mean alone cannot separate those effects.
Second: "The treatment worked, so it explains the gap." Ask who received the treatment and who they were compared with. Did the trial test the gap itself? A treatment effect in one selected group does not settle that question.
Third: "An adjusted link proves cause." Ask how the study dealt with other possible causes. Were people assigned to groups by chance? If not, what other design features support a causal claim? Including measured factors in a model leaves room for those it missed.
For your own relationship, ask what each partner wants and how you discuss it. You do not need to decide which group mean is right to hear each other. That conversation is not a clinical assessment.
Next in this series: Sexual violence risk factors. Browse the full Insights library.
Editor's note: This article is for general information only and is not a substitute for professional advice. If you or someone you know is affected by sexual violence, please reach out to a qualified professional or a local support service.
Key takeaways
- Group averages describe groups. A moderate average gap can sit beside substantial overlap between people.
- A treatment trial and a population comparison answer different questions. One does not settle the other.
- Observational links can survive adjustment without proving cause.
- Desire, arousal, consent, and behavior are separate. Clear, ongoing agreement is the standard.
Frequently asked questions
References
Conley, T. D., & Yang, T. A. (2024). Can sociocultural and contextual factors explain gender differences in sex drive? A response to Frankenbach et al. (2022). Psychological Bulletin, 150(8), 1004-1010. https://doi.org/10.1037/bul0000412
Frankenbach, J., Weber, M., Loschelder, D. D., Kilger, H., & Friese, M. (2022). Sex drive: Theoretical conceptualization and meta-analytic review of gender differences. Psychological Bulletin, 148(9-10), 621-661. https://doi.org/10.1037/bul0000366
Frankenbach, J., Weber, M., Loschelder, D. D., Kilger, H., & Friese, M. (2024). Gender differences in sex drive: Reply to Conley and Yang (2024). Psychological Bulletin, 150(8), 1011-1019. https://doi.org/10.1037/bul0000444
RAINN. (2026, May 31). Consent 101: Respect, boundaries, & building trust.
Snyder, P. J., Bhasin, S., Cunningham, G. R., Matsumoto, A. M., Stephens-Shields, A. J., Cauley, J. A., Gill, T. M., Barrett-Connor, E., Swerdloff, R. S., Wang, C., Ensrud, K. E., Lewis, C. E., Farrar, J. T., Cella, D., Rosen, R. C., Pahor, M., Crandall, J. P., Molitch, M. E., Cifelli, D., ... Ellenberg, S. S. (2016). Effects of testosterone treatment in older men. The New England Journal of Medicine, 374(7), 611-624. https://doi.org/10.1056/NEJMoa1506119
Wright, P. J., Paul, B., & Herbenick, D. (2021). Pornography, impersonal sex, and sexual aggression: A test of the confluence model in a national probability sample of men in the U.S. Aggressive Behavior, 47(5), 593-602. https://doi.org/10.1002/ab.21978
This article is for informational and educational purposes only and does not constitute financial, legal, tax, medical, or professional advice. Individual results vary.