RN
Somebody, somewhere, decided that the answer to this is a number, and the number is once a week. It gets repeated in magazine copy, in relationship threads, in the comments under every article on the subject, usually with the confidence of a physical constant.
It is worth knowing where that figure came from, because the study behind it is real, it is better than most of what surrounds it, and it does not mean what the repetition has turned it into.
The number and its actual footnotes
The source most often being half-remembered is a 2015 paper by Amy Muise, Ulrich Schimmack and Emily Impett in Social Psychological and Personality Science. Working across large survey datasets, they found that self-reported well-being rose with sexual frequency up to roughly weekly, and then flattened — couples reporting more than that were not, on average, reporting proportionally happier lives.
Four things about that finding get dropped on the way to your feed.
It described people in established relationships. It was not a result about single people, and the authors said so.
It was correlational. A design like that cannot tell you whether the sex produced the well-being, the well-being produced the sex, or a third thing — health, free time, an absence of crisis — produced both. The honest reading is that people whose lives are going reasonably well tend to have more sex and also tend to report feeling better, which is not a scandal but is also not an instruction.
It was an average across a very large group. Averages of that kind conceal enormous spread. A population can average weekly while containing a great many people at zero and a great many at daily, and the average describes none of them.
And it was a plateau, not a peak. The finding was that more stopped adding, not that more subtracted.
Why the underlying numbers are softer than they look
Nearly everything published about how often people have sex traces back to self-report surveys — in the US, principally the General Social Survey and the National Survey of Sexual Health and Behavior; in the UK, Natsal. There is no other practical way to collect the data, and there is also no way to collect it cleanly.
People are being asked to recall private behaviour over a long window, to a stranger, about a subject with strong social scripts attached. Recall degrades. Social desirability pushes answers toward whatever the respondent thinks is expected, and it pushes in different directions for different groups. Question wording matters: what counts as sex is not defined identically across surveys, which quietly makes the datasets less comparable than the round numbers imply.
Then there is what the samples cover. Averages are usually reported for adults broadly, mixing partnered and unpartnered people, which drags the figure toward whatever the partnering rate happens to be in that cohort — a fact about relationship formation being read as a fact about libido.
Cohort trends run into the same trap. Analyses of General Social Survey data led by Jean Twenge found reported adult sexual frequency in the US declining across recent decades. That result has held up reasonably well and is worth knowing. It is also compatible with several different explanations — later partnering, more people living alone, changes in what respondents are willing to report — and it says nothing about whether any individual is doing something wrong.
None of this means the research is bad. It means it is population-level research, doing what population-level research does, and it was never designed to adjudicate your particular month.
What clinicians actually attend to
Ask people who treat sexual difficulty for a living and frequency is rarely the presenting variable. The one that is: the gap between partners.
Desire discrepancy — one person wanting sex noticeably more often than the other — is the pattern that generates distress, and it does so at any absolute frequency. Two people who both want it twice a month have no problem. Two people who want it twice a month and four times a week have a real one, and it does not improve by either of them consulting a national average.
The second thing is a model of desire most people have never encountered. The intuitive picture is spontaneous desire: wanting appears unprompted, and you act on it. Rosemary Basson's work described a responsive pattern instead, common particularly though not exclusively in women, where desire arrives after arousal has begun rather than before it. Under that pattern, waiting to feel spontaneously in the mood before initiating anything can mean waiting indefinitely, and the resulting "I have no sex drive" conclusion is often a mismatch between a real pattern and an assumed one.
Neither of those is measured by a frequency count, which is roughly why frequency counts do not predict who is unhappy.
What actually moves it
Sexual frequency is downstream of a long list of unromantic things: sleep, workload, illness, chronic pain, alcohol, depression and anxiety, relationship conflict, small children, and medication — SSRIs in particular are well documented as affecting desire and orgasm, and so are several hormonal contraceptives and blood-pressure drugs.
That list is why frequency moves around over years without anything being broken. It also contains the reason to see somebody: a marked, sustained change from your own baseline, especially one arriving alongside fatigue, mood change, pain during sex, or a new prescription, is worth raising with a doctor. Thyroid disorder, low testosterone, depression and drug side effects are all identifiable and mostly manageable, and none of them is diagnosable from an article. What does not warrant a clinical visit is being a different number from a survey.
The comparison you cannot actually make
The last problem with the whole exercise is that the comparison group is imaginary.
Friends' accounts of their sex lives are edited. Survey averages describe a population, not a person. And anything you have watched is a filmed performance, cut for pace, made by people at work — a point worth holding onto if the porn addiction versus high libido question has ever occurred to you, and the reason talking to a partner directly outperforms inference from any of those sources.
There is a defensible answer to "how often should you have sex," and it is unsatisfying: often enough that neither person in it feels the frequency as a problem. That is a question with one relevant respondent — or two — and none of them are in a dataset.