There's No Such Thing as the 'Most Unfaithful Country' — Here's What the Data Actually Shows

Search "most unfaithful country" and you'll get a list with a percentage next to a flag. Somewhere a country tops it, somewhere else sits at the bottom, and the whole thing reads like a fact you could repeat at dinner. It isn't one. Before you get to why, here's the single number that should make you distrust every list like it, including any this article could have written:

In one national survey, the same women were asked about infidelity two different ways in the same year — and reported it at 1.08% face-to-face versus 6.13% by computer. Same country. Same respondents. A six-fold difference produced by nothing except how the question arrived.

That's not a footnote to the "most unfaithful countries" question. It's the answer to it.

The Study That Should End the Debate

The number above comes from Whisman and Snyder (2007, *Journal of Family Psychology*, 21(2), 147–154, DOI 10.1037/0893-3200.21.2.147). They worked with a population-based US sample of 4,884 married women and asked about past-year sexual infidelity in two ways: a face-to-face interview with a live interviewer, and a computer-assisted self-interview (ACASI), where the respondent reads the question on a screen and answers privately, with no interviewer hearing the response.

Face-to-face: 1.08% reported infidelity in the past year. Computer: 6.13%. Same women, same period, same underlying behavior — whatever it actually was. The only thing that changed was the delivery mechanism for the question.

If method alone can move a rate by 6x within one country, comparing "Country A's rate" to "Country B's rate" only tells you something real if both countries used the same method, on comparably drawn samples, asking about the same behavior. That almost never happens. Most rankings you'll find online splice together numbers from different surveys, different decades, different question wordings, and different research traditions, then present the result as if it were one measurement.

A Meta-Analysis Says the Same Thing, at Much Larger Scale

This isn't a one-off result. Warach and colleagues (2024, *Personal Relationships*, DOI 10.1111/pere.12571) pooled 305 studies covering 508,241 respondents across 47 countries — a systematic review and meta-analysis of romantic infidelity prevalence and what actually moves it. Their conclusion: how infidelity was measured (in-person versus anonymous, broad definition versus narrow, retrospective recall versus daily diary) accounted for far more of the variation between studies than the country the sample came from did.

Put plainly: if you took every published infidelity study and tried to explain why the numbers differ, "which country" is a weaker predictor than "which survey design." That finding alone should retire the ranking format. A list that orders countries by reported rate is, structurally, mostly ordering them by which studies happened to survey them and how.

There Isn't Even an Agreed Definition to Rank By

The problem goes one layer deeper than measurement error. Blow and Hartnett published a two-part review of the infidelity literature in the *Journal of Marital and Family Therapy* in 2005 — a methodological review (31(2), 183–216, DOI 10.1111/j.1752-0606.2005.tb01555.x) and a substantive review (31(2), 217–233, DOI 10.1111/j.1752-0606.2005.tb01556.x). Across the field, they found no consensus definition of what counts as infidelity. Some studies count only penetrative sex. Others count any sexual contact, emotional affairs, sustained secret texting, or a single kiss. Some ask about "cheating"; others ask about "extramarital sex," which respondents interpret differently.

A ranking assumes every country's number answers the same question. Given Blow and Hartnett's finding, it usually doesn't — one country's figure might count only intercourse, another's might include emotional affairs, and the table gives you no way to tell which is which.

What You Can Actually Trust: Trends, Not Tables

None of this means infidelity is unmeasurable — it means the *comparison across countries* is the weak link, not measurement itself. Two things hold up:

Within-country trends from one consistent instrument. The US General Social Survey has run since 1972, though its specific extramarital-sex question (variable EVSTRAY: "Have you ever had sex with someone other than your husband or wife while you were married?") was added later, in 1991, and has been asked in every wave since. Because the same question goes to the same country's population repeatedly over decades, a change in the GSS's own number over time is at least measuring the same thing against itself. That's a different — and far more defensible — kind of claim than "the US rate is higher than France's rate," which reintroduces every cross-survey problem above.

Validated predictors, replicated across studies. Instead of asking which country cheats more, better-supported research asks what predicts infidelity within a relationship, regardless of where it happens. Two of the most consistent findings: relationship dissatisfaction and low commitment. Treas and Giesen (2000, *Journal of Marriage and Family*, 62(1), 48–60, N=2,598 ever-married or cohabiting US adults) linked lower relationship quality to higher odds of infidelity. Wiederman (1997, *Journal of Sex Research*, 34(2), 167–174, DOI 10.1080/00224499709551881, N=2,172 from the 1994 GSS) found comparable patterns tying extramarital sex to relationship-quality measures rather than to demographic labels. Neither paper ranks nations. Both point at the relationship itself.

The Study People Misquote to Rank Countries Anyway

One more citation shows up constantly in "unfaithful nations" content, almost always used wrong. Betzig (1989, *Current Anthropology*, 30(5), 654–676, DOI 10.1086/203798) coded ethnographic records from 160 traditional societies for stated causes of divorce. Out of 43 coded causes, infidelity was the single most commonly cited reason people gave for ending a marriage.

That is a real, striking finding — and it is not a cheating rate. Betzig counted *stated reasons for divorce*, not incidents of infidelity. A society could have relatively low actual infidelity but treat it as unforgivable grounds whenever it happens, making it a frequently *cited* cause. Another society could have more infidelity but tolerate or quietly absorb it, so it rarely gets named as the reason a marriage ended. "Most commonly cited cause of divorce" and "highest rate of cheating" are different measurements of different things, and Betzig's data supports only the first one.

Method

Every specific claim above is drawn from the peer-reviewed sources named inline — Whisman & Snyder (2007), Warach et al. (2024), Blow & Hartnett (2005a, 2005b), Treas & Giesen (2000), Wiederman (1997), and Betzig (1989) — each linked to its DOI so you can check the original rather than this summary. None of these papers rank countries by infidelity, and this article doesn't either.

This piece is part of a research series in which every claim carries its source and the underlying data is published alongside the article. In that spirit, elsewhere in this series we publish our own raw data rather than just asserting conclusions from it. It isn't the source for anything claimed in this article, but it's built the same way: pulled from a documented public API, with the collection method and its limits stated up front. If you want to see that approach applied to a different question (search interest in breakup- and divorce-related topics, by month and by language edition of Wikipedia), the datasets are open: relationship-seasonality.json, relationship-seasonality.csv, and relationship-vocabulary.json, which tracks the rise of terms like limerence and love bombing on Wikipedia over the past decade. For more details, see our guide on follower tracker guide. For more details, see our guide on see who unfollowed you.

FAQ

So which country actually is the most unfaithful?

There's no sound answer to that question with current data, and that's the point of this article rather than a dodge. Warach et al. (2024) found that survey methodology explains more of the variation in reported infidelity than country does, across 508,241 respondents in 47 countries. A ranking built from mismatched surveys isn't measuring what it claims to.

Why does asking the same people by computer versus in person change the answer so much?

Admitting infidelity to a live interviewer carries a social cost that answering privately on a screen doesn't. Whisman and Snyder (2007) documented the size of that gap directly: 1.08% versus 6.13% in the same national sample of American women. This kind of self-report gap — sometimes called social desirability bias — applies to any sensitive-behavior survey, not just this one.

Is there any reliable predictor of infidelity at all?

Yes — within a relationship, not across nations. Relationship dissatisfaction and lower commitment are the most consistently replicated predictors, found separately by Treas and Giesen (2000, N=2,598) and Wiederman (1997, N=2,172). Neither is a nationality effect; both are about the state of the specific relationship.

What about the claim that infidelity is "the top cause of divorce" in most cultures?

That comes from Betzig (1989), who coded 160 traditional societies and found infidelity was the most frequently *cited* cause of marital dissolution, out of 43 coded causes. That measures how often people name infidelity as the reason a marriage ended — a cultural and legal framing — not how often infidelity itself occurs. The two get conflated constantly, but Betzig's data only supports the first claim.