Three decades of infidelity data reveal unexpected patterns

The core GSS figures (20% men, 13% women) primarily capture sexual infidelity as defined by participants. Broader definitions that include emotional affairs, texting relationships, and non-physical intimate connections produce higher prevalence estimates, commonly cited as up to 25% overall. The definitional boundary varies by study and by individual, which is one reason infidelity prevalence data varies significantly across published sources.

No. Research on relationship outcomes following disclosed infidelity consistently finds that the majority of couples do not separate immediately following discovery. Gordon et al. 2004 (Journal of Consulting and Clinical Psychology) found that about 65% of couples remained together at least initially following disclosure. Longer-term outcomes are more mixed. The decision to remain or leave following infidelity is complex and highly individual; the data does not support either "always leave" or "always stay" as a statistically dominant outcome.

Yes. Gottman-method couples therapy and Emotionally Focused Therapy (EFT) both have documented evidence of effectiveness in couples processing infidelity. Gordon et al. 2004 found that couples who received targeted intervention showed significantly better relationship satisfaction outcomes than those who did not. The process is often described as addressing the attachment injury rather than the specific act, rebuilding trust and security rather than litigating events. Not all couples choose to pursue therapy, and neither staying nor leaving requires therapeutic intervention as a condition.

Physical infidelity involves sexual contact outside a committed relationship. Emotional infidelity involves romantic attachment, sustained intimate communication, or deep emotional dependency with someone outside the relationship, without necessarily involving physical contact. Studies find that men and women can differ in which type they find more threatening, though the finding has been contested in more recent research. More practically relevant: definitions of infidelity vary significantly between individuals in the same relationship, and mismatched definitions are a documented source of conflict separate from any actual behaviour.

GSS data shows men consistently self-report higher rates of lifetime infidelity: approximately 20% versus 13% for women across the most recent multi-decade aggregation. However, the gender gap has narrowed significantly over time. Among adults under 45, the gap is smaller than in older cohorts, with some studies finding it near parity. The difference in reported rates partly reflects social desirability bias: men in some social contexts face less stigma for disclosure, women face more. Both numbers are likely underestimates of actual prevalence. The full GSS infidelity data, including trends over time and demographic breakdowns, is compiled on the relationship statistics page.

Infidelity research distinguishes between "reactive" affairs (responses to relationship deficits) and "opportunistic" affairs (where relationship quality is not the primary driver). Glass and Wright 1992 found that a substantial proportion of people who reported affairs also reported being satisfied or very satisfied with their primary relationship. Cited motivations include novelty-seeking, desire for validation, circumstantial opportunity, and the Alter and Hershfield 2014 finding that milestone ages produce identity-seeking behaviour. Attachment style is a significant factor: individuals with anxious attachment patterns (characterised by fear of abandonment and hypervigilance about partner availability) and those with avoidant attachment patterns (characterised by discomfort with intimacy and emotional distance) both show elevated infidelity rates relative to securely attached individuals, through different pathways. Relationship dissatisfaction is a predictor of infidelity, but it is neither necessary nor sufficient.

Yes. Research by Knopp et al. 2017 (Archives of Sexual Behavior) found that people who had been unfaithful in one relationship were three times more likely to be unfaithful in a subsequent relationship. The effect held after controlling for relationship satisfaction. This suggests a partially stable individual disposition toward infidelity, likely reflecting a combination of personality traits, attitudes toward commitment, and the lowering of psychological barriers after the first instance. Prior infidelity is one of the stronger individual-level predictors in the literature.

Discovery of a partner's infidelity is consistently associated with elevated rates of anxiety, depression, and post-traumatic stress symptoms in the betrayed partner. Research using PTSD diagnostic criteria has found that a clinically significant proportion of betrayed partners meet criteria for post-traumatic stress in the immediate aftermath. The perpetrator also experiences psychological costs: guilt, shame, and relationship anxiety are documented even among those who do not disclose. Recovery trajectories vary substantially and are influenced by disclosure quality, partner response, and whether the affair has ended.

Infidelity research identifies several recurring affair types: opportunistic (situationally-driven, low emotional investment), romantic (emotional attachment without long-term intent), conflicted romantic (genuine attachment to both partners simultaneously), and exit affairs (a vehicle for leaving the primary relationship). Opportunistic and conflicted romantic types appear most frequently in survey data. Exit affairs are the minority despite featuring prominently in cultural narratives. The type of affair influences recovery prospects: romantic affairs with emotional attachment are associated with more severe relationship damage than opportunistic encounters.

It depends on how the couple has defined their relationship's boundaries. Surveys consistently find that most people classify sustained emotional online relationships, explicit video calls, and sexting with someone outside the relationship as infidelity. Dating app use, social media messaging that includes flirtation or emotional intimacy, and sustained contact with an ex through private channels are also frequently cited as boundary-violating behaviours in population surveys, particularly among younger adults who have grown up with digital social environments as a normal feature of social life. Viewing pornography is classified as infidelity by a significant minority. The cultural conversation around digital infidelity is relatively recent and population norms are still forming. What matters clinically is not the cultural average but whether a behaviour has violated the specific agreements of a particular relationship.

Non-monogamous relationships (consensual non-monogamy, or CNM) address infidelity by restructuring the agreement rather than preventing the behaviour. Research from Rubel and Bogaert 2015 (Journal of Sex Research) found no evidence that CNM relationships have higher rates of dissatisfaction or lower trust than monogamous ones, though the populations differ. CNM is not a guarantee against boundary violations: partners in CNM arrangements also report experiences of agreement breaches. The key difference is that violations in CNM involve breaking negotiated agreements rather than a universal monogamy norm.

The research on child outcomes focuses more on parental conflict and household instability than on infidelity per se. When disclosure leads to sustained conflict, disrupted parenting capacity, or separation, child wellbeing can be affected through those pathways. Infidelity that is managed without prolonged conflict or household disruption shows less direct child impact in the literature. The quality and stability of parenting following disclosure matters more to child outcomes than the infidelity event itself.