Infidelity Trends: 1990–2025

Infidelity trends data has held steady in aggregate: IFS researcher Wendy Wang’s 2023 GSS analysis found 20% of ever-married men and 13% of ever-married women report sex outside marriage — but those headline figures mask dramatic generational shifts between 1990 and the 2020s. This post traces rates by decade, gender, and age cohort using primary-source General Social Survey data.

What Counts as Infidelity in Longitudinal Research

In longitudinal research, “infidelity” typically refers to sex with someone other than one’s spouse while married — the precise framing used in the U.S. General Social Survey’s EVSTRAY question, which has been asked of nationally representative samples since the early 1990s. This definition excludes emotional-only affairs and cohabiting (non-married) partnerships, meaning published trend data captures a narrower slice of infidelity than the word implies in everyday language. Researchers at the Institute for Family Studies 2021 describe this as a deliberate methodological choice — not a flaw — suited to cross-decade comparison.

The gap between the behavioural measure and the public’s own definition is substantial. The IFS’s 2019 iFidelity Survey, conducted by YouGov with 2,000 nationally representative U.S. adults, found that 76% of respondents considered a secret in-person emotional relationship to constitute infidelity, and 72% said the same about a secret online emotional relationship — yet these forms are entirely absent from the GSS’s standard measure, per IFS 2020.

The same IFS analysis showed that traditional survey measures counting only physical sex capture fewer than half of all affairs people self-report. Breaking down the ever-married population: 7% reported a strictly emotional affair, 5% a sexual-only affair, and 10% an affair that was both sexual and emotional. The majority of the “gap” between the GSS’s narrow measure and real-world prevalence is occupied by emotional affairs that leave no trace in standard longitudinal datasets. For partners trying to recognize the patterns themselves, see our blog on signs of an emotional affair.

For an overview of current headline rates across all affair types, see the infidelity statistics pillar hub.

What the GSS Measures — and What It Misses

The General Social Survey is the most widely used primary source for long-run U.S. infidelity trend data, but its design carries specific constraints that affect how the numbers should be read:

• Single behavioural question, fielded intermittently. The GSS has tracked infidelity since the early 1990s via a question asking whether respondents have “had sex with someone other than their spouse while married.” Because the GSS is an omnibus household survey, the infidelity module is not fielded in every annual wave — limiting the granularity of decade-level trend analysis, per IFS 2023.
• Face-to-face interview method produces systematic underestimates. A 2007 study by Whisman and Snyder published in the Journal of Family Psychology (N=4,884 married women) found annual infidelity prevalence of 1.08% via face-to-face interview, compared with 6.13% when the same population used a computer-assisted self-interview — a nearly six-fold difference. This means the GSS’s in-person methodology likely produces systematic underestimates of true rates, per NCBI PubMed 2007.
• Remarriage ambiguity. The GSS asks about any spouse, not specifically the current one. A portion of “cheaters” in the GSS data may have had their affair in a previous marriage — a limitation noted explicitly in IFS analyses. As IFS states: “we can’t tell from the data whether or not the person who cheated is still married to the spouse he or she cheated on,” per IFS 2023.
• Physical-sex-only scope. The GSS question captures penetrative sexual activity outside marriage and excludes emotional affairs, online infidelity, and non-penetrative sexual contact — meaning it measures a subset of what most Americans identify as cheating.
• No cohabiting relationships. Unmarried couples in long-term partnerships are outside the GSS’s scope entirely. As cohabitation has become more common across all age cohorts since the 1990s, an increasing share of romantic infidelity simply falls outside the survey’s frame.

How Infidelity Rates Have Changed by Decade: 1990s to 2020s

Across three decades of GSS data analyzed by the Institute for Family Studies, the overall infidelity rate has not increased dramatically — but the age profile of who cheats has shifted substantially.

The 1990s peak figures — 31% for men aged 50–59 and 18% for women aged 40–49 — were concentrated in the cohorts born in the 1940s and 1950s who came of age during the sexual revolution, per IFS 2023. By the 2000s, GSS trend data tracked by IFS shows the peak infidelity age had shifted upward: men aged 60–69 recorded the highest rate at 29%, and women aged 50–59 reached 17%. The same high-infidelity cohort was aging through the data rather than a new wave of younger cheating emerging.

By 2016, the overall U.S. infidelity rate had held “relatively steady at around 16 percent” according to IFS researcher Nicholas Wolfinger’s 2017 GSS analysis — but with a widening age divide: adults over 55 reported infidelity at approximately 20%, compared to 14% for those under 55.

In the most recent GSS wave analyzed by IFS, the lifetime rate for ever-married men stood at 20% overall, while for women it was 13%. The striking finding for the most recent cohort: among adults aged 18–29, the male and female rates had nearly converged to 10% (men) and 11% (women), with young women slightly exceeding young men.

The Gender Gap Over Time: Men vs. Women, 1990–2025

Across all age groups in aggregate GSS data, men have consistently been more likely than women to report infidelity. The overall gap has persisted while narrowing sharply among younger cohorts.

Among adults aged 80 and older, the gender gap in infidelity grew from 5 percentage points in the 1990s to 12 percentage points by the 2000s, consistent with men in that cohort having had greater access to opportunity and, from 1998 onward, erectile dysfunction medications — per IFS 2023.

The IFS’s 2019 iFidelity Survey, conducted by YouGov with 1,282 ever-married adults, used a narrower physical-only definition and found 20% of ever-married men and 10% of ever-married women had cheated — a 10-point gap slightly wider than the GSS gap, likely because the iFidelity question excluded emotional affairs, which women report disproportionately, per IFS 2021.

IFS’s emotional infidelity analysis of the same iFidelity Survey adds further context: men comprised 75% of sex-only affair participants and 56% of combined sexual-and-emotional affair participants, while women accounted for 56% of emotional-only affair participants — reinforcing that the measured gender gap narrows significantly when emotional infidelity is included in the definition, per IFS 2020.

For current headline rates broken down by gender and additional demographic cuts, see Infidelity Statistics: Who Cheats More, Men or Women?. For how those rates are shifting onto digital channels — dating apps, social media, and AI companions — see the Digital Infidelity Report.

How Age Cohorts Shifted: Who Cheats Most Has Changed

The most important structural change in U.S. infidelity data between 1990 and the 2020s is generational: a single birth cohort — Americans born between 1940 and 1959 — has dominated the infidelity statistics throughout the entire measurement period, and their aging explains most of what looks like trend movement.

IFS researcher Nicholas Wolfinger’s 2017 analysis of 30 years of GSS data identified Americans born between 1940 and 1959 — the baby boomers — as reporting the highest lifetime rates of extramarital sex, calling them “the first generations to come of age during the sexual revolution.” By 2016, adults over 55 reported infidelity at roughly 20%, compared to 14% for those under 55.

Among ever-married adults aged 18–29 in the most recent GSS data, IFS 2023 found the infidelity rate was 10% for men and 11% for women — effectively equal, and markedly lower than the 20–31% peak rates recorded for the same age group in the cohorts born two to three decades earlier.

Attitude data shows a parallel shift. GSS disapproval of marital infidelity held between 77% and 83% from 1988 to 2008, then dropped nearly five points between 2008 and 2010, with further fluctuations through 2014 — a pattern IFS 2015 tracked as consistent across all age groups, debunking the claim that millennials alone drove any attitudinal shift.

In 2014, GSS respondents aged 18–33 were the age group least likely to condemn infidelity unequivocally — by about four percentage points compared to older cohorts. However, IFS notes this difference was within the survey’s ±3–4 point margin of error, making the millennial-specific claim statistically fragile. A separate analysis by IFS 2016 using a 2014 study by Paul Hemez (National Center for Family and Marriage Research) found 75% of millennials (then aged 18–33) disapproved of infidelity — the lowest disapproval rate of any age cohort, when national disapproval overall had fallen to 79% from a peak of 84% in 2008.

Why Infidelity Rates Are Hard to Compare Across Countries

Cross-national infidelity comparisons are methodologically unreliable, and the reasons go beyond cultural differences. Three distinct design factors make published rates from different countries non-comparable without substantial qualification.

The first factor is survey method. A peer-reviewed study by Whisman and Snyder 2007, published in the Journal of Family Psychology, demonstrated that the choice of interview method alone produces a nearly six-fold difference in measured infidelity rates — 1.08% via face-to-face interview versus 6.13% via anonymous computer-assisted self-interview among the same population of married women. When countries use different survey modes, their published rates are not directly comparable even before cultural factors enter the picture.

The second factor is definition. The IFS’s iFidelity Survey used a deliberately narrow definition — “physical sexual activity with someone other than their spouse, and without their spouses’ knowledge and consent” — which IFS explicitly acknowledges is narrower than some previous research. Surveys in other countries that include emotional affairs, one-night-stands, or cohabiting relationships will yield different baselines by design, not just by culture, per IFS 2021.

The third factor is reference population. YouGov’s 2022 U.S. survey (n=2,000) found that 33% of Americans who had ever been in a monogamous relationship reported having cheated — physically, emotionally, or both — compared to just 13% who reported cheating on a current married partner specifically. This illustrates how the reference population (“ever in any relationship” vs. “currently married”) produces dramatically different headline figures even within the same country and survey, let alone across national boundaries.

For these reasons, country-by-country infidelity rate tables circulating in media should be treated with considerable skepticism. The most defensible cross-national comparison available from whitelisted primary sources is between U.S. subgroup rates (GSS, IFS) and U.S. subgroup rates measured by alternative methodologies — the comparison above from YouGov being a case in point.

What These Trends Mean for Couples — and for Investigation

The data carries practical implications beyond academic interest. Two findings stand out for anyone assessing whether a suspected infidelity is likely to continue or recur.

First, serial infidelity is a documented pattern. A longitudinal study by Knopp et al. 2017 — published in the Archives of Sexual Behavior — tracked 484 adults through two consecutive relationships and found that 45% of those who cheated in their first relationship also cheated in their second, compared to 18% of those without prior infidelity. Those who had cheated were three times more likely to cheat again. Prior infidelity is the strongest single predictor of future infidelity in the peer-reviewed literature. For a practical breakdown of the behavioural patterns reported by clients seeking confirmation, see our guide on the signs of a cheating partner.

Second, confirmed infidelity has a strong correlation with marital dissolution. Among people who reported cheating in IFS’s 2023 GSS analysis, 40% were currently divorced or separated, compared to 17% among those who remained faithful — meaning confirmed infidelity roughly doubles the probability of marital dissolution. Where suspicion has already tipped toward separation, evidence-gathering for divorce investigations follows a different admissibility workflow than pure infidelity confirmation.

As licensed PIs operating across major Canadian cities, we see these patterns reflected in our client enquiries. The age and gender cohort shifts documented by IFS are visible in the demographics of who seeks investigation: clients seeking evidence of a spouse’s infidelity are increasingly spread across age groups rather than concentrated in the older cohorts that historical data would have predicted. The convergence in younger cohort rates also means that assumptions based on gender alone — that men are more likely to be cheating — are less reliable for adults under 35 than they were a generation ago.

If you suspect a partner’s infidelity and want to understand your investigative options, the cheating spouses investigation service outlines how licensed investigators gather admissible evidence. For a broader view of the infidelity research our analysts track, see the infidelity statistics pillar hub and the full infidelity statistics breakdown at Infidelity Statistics: Who Cheats More.