Couples Therapy

Abstract

Couples therapy is a form of psychotherapy which MeSH classifies under group psychotherapy: the treatment of two partners together, with the relationship, not either individual, as the unit of care. It began in the 1960s and 1970s as behavioral marital therapy, teaching couples to exchange rewarding behaviors and to communicate and solve problems better, and it matured through two influential turns — Andrew Christensen's integrative behavioral couple therapy, which added emotional acceptance to behavior change, and Susan Johnson's emotionally focused therapy, which recast distress as a disruption of the attachment bond. John Gottman's observational research established that specific negative-affect patterns, not the presence of conflict, predict which marriages dissolve. Meta-analyses now place couple therapy among the more effective psychotherapies, with moderate-to-large effects that roughly half of couples maintain over five years, and technology-assisted delivery has begun to widen its reach.

Keywords: couples therapy, integrative behavioral couple therapy, emotionally focused therapy, marital distress, attachment

What Couples Therapy Is

Couples therapy is the treatment of relationship distress by working with both partners together, organized around the premise that the object of treatment is the relationship itself — the recurring patterns of interaction the two people generate — rather than a symptom lodged in one of them. A partner may arrive convinced the problem is the other person, or that it is their own failing; the therapist's reframing move is to locate the difficulty in the cycle that the couple enacts, a pattern that neither partner intends and both maintain. Because the unit of care is a two-person system, the therapy is conducted with both in the room, and change is sought in how they respond to each other rather than in either alone (Gurman & Fraenkel, 2002).

This systemic framing has a specific empirical warrant. Decades of research on what actually distinguishes distressed from satisfied couples found that the discriminating variable is not the amount of conflict but its conduct: satisfied couples disagree as often, but distressed couples fall into self-amplifying sequences of negative affect that each partner's response feeds. The therapeutic target follows directly — interrupt the sequence, change what each partner does at the moments the cycle turns, and the relationship can recover even when the presenting disagreements do not disappear (Snyder et al., 2006).

Couples therapy is distinguished from individual therapy delivered to two people by exactly this focus on the interaction. The therapist attends less to each partner's internal history than to what happens between them in the room — how a bid for closeness is met, how a complaint escalates, how a withdrawal provokes pursuit — because that observable, present-tense exchange is both the disorder and the site of its repair (Gurman & Fraenkel, 2002). One such exchange recurs so reliably that it has become the field's paradigm case: the demand-withdraw sequence, in which one partner presses for change or closeness while the other retreats, an asymmetry whose structure and gendered correlates Christensen and Heavey mapped in the study that made it a named construct (Christensen & Heavey, 1990).

Origins: From Behavior Exchange to Emotion and Acceptance

Couples therapy in its modern, empirically studied form began with the application of learning principles to marriage. Behavioral marital therapy, developed in the 1960s and 1970s, treated relationship distress as a deficit in the exchange of rewarding behavior and in the skills of communication and problem-solving; its interventions taught couples to increase positive exchanges and to negotiate conflict through structured, teachable skills. By the late 1990s a systematic review of the empirical literature could identify behavioral marital therapy as the best-established, empirically supported treatment for marital distress, giving the young field a firm evidential anchor (Baucom et al., 1998).

Alongside the treatment research ran a programme of basic observation that reshaped how distress was understood. John Gottman and Robert Levenson followed couples longitudinally, coding their interactions behavior by behavior and recording their physiology, and found that a specific profile of negative-affect exchange — not conflict as such — predicted which marriages would later dissolve. Distress, on this account, is a process with an observable signature, and the discovery gave couples therapy a science of what it was trying to change (Gottman & Levenson, 1992).

The behavioral programme then split along two lines that still organize the field. One line questioned whether teaching couples to change each other was always the right aim: some differences are stable and non-negotiable, and pushing for change can entrench the conflict. That insight became the acceptance turn, formalized in integrative behavioral couple therapy. The other line questioned the behavioral account of emotion itself. In the first controlled outcome study of what became emotionally focused therapy, Susan Johnson and Leslie Greenberg contrasted an experiential, emotion-focused intervention with a behavioral problem-solving one and found the emotion-focused approach produced greater gains — evidence that engaging partners' underlying emotional experience, rather than their behavior alone, could be the more powerful lever (Johnson & Greenberg, 1985).

Types of Couples Therapy

MeSH files couples therapy beneath group psychotherapy — a two-person dyad being, for indexing purposes, the smallest group — and gives it one narrower descriptor, its single direct child in the thesaurus. The distinction the classification draws is one of framing rather than of competing schools: whether the two partners are treated as a couple, a bond to be strengthened, or as a marriage, a formalized relationship whose conflicts are the object of care. The MeSH tree is an indexing scheme for retrieving literature, not a theory of the therapy, and in practice the two headings overlap almost entirely — most of what is published under one is relevant to the other.

Marital therapy is the older heading, dating to a period when the treated couples were assumed to be married; couples therapy is the broader contemporary term, covering unmarried, cohabiting, and same-sex partners on the same relational principles. The narrower descriptor is noted here for completeness and is not yet a pillar topic in its own right. What unites the two is the premise that treatment happens through the relationship between two partners rather than within either one (Gurman & Fraenkel, 2002).

The Major Models

Three models dominate the empirically supported practice of couples therapy, and each locates the engine of change in a different place. Traditional behavioral couple therapy locates it in behavior: distress is a shortage of rewarding exchange and a deficit of communication skill, and the treatment increases the first and teaches the second. It works, but its gains are strongest where the couple's difficulties are genuinely negotiable, and a substantial fraction of couples relapse — the limitation that motivated the two later models (Baucom et al., 1998).

Integrative behavioral couple therapy (IBCT), developed by Andrew Christensen and Neil Jacobson, keeps behavior change but adds emotional acceptance as a co-equal aim. Its central recognition is that some differences between partners are stable and cannot be negotiated away, and that the couple's suffering often comes less from the difference itself than from the polarized struggle over it. IBCT therefore works to help each partner accept what cannot be changed — to hear the vulnerable feeling beneath the other's position — while still pursuing change where change is possible. In the pivotal randomized trial comparing IBCT with traditional behavioral couple therapy in more than a hundred chronically and severely distressed couples, both treatments helped, with IBCT showing a distinct pattern of gains (Christensen et al., 2004). The acceptance-and-change model and its dissemination, including its adaptation for web-based delivery, are laid out in a contemporary account of the approach (Roddy et al., 2016).

Emotionally focused therapy (EFT), developed principally by Susan Johnson, locates the engine of change in the attachment bond. Drawing on attachment theory, it treats relationship distress as a form of separation distress — a protest against the felt loss of a secure connection — and its recurring pattern of one partner pursuing while the other withdraws as a cycle of insecure attachment. The therapy works to de-escalate that cycle and then to restructure the bond through emotionally engaged encounters in which partners can express attachment needs and respond to each other's. The approach was defined as a distinct, researchable treatment in a status-and-challenges review that reported large effect sizes and set its research agenda (Johnson et al., 1999), and its attachment-based outcome and process literature has since been synthesized in a modern review (Wiebe & Johnson, 2016).

Couples Therapy in Motion

The three demonstrations below make the models manipulable. The first builds the IBCT dialectic of change and acceptance into a model of distress, showing when accepting a difference beats trying to change it. The second animates the demand-withdraw cycle, showing how a small repair rate decides whether conflict escalates or settles. The third turns to outcome, translating a meta-analytic effect size into the proportion of couples the therapy helps.

Rate an issue’s severity, how hard the couple pushes to change it, how changeable it actually is, and how far the partners accept it. Change effort reduces the issue only as far as the issue is changeable, giving a residual I′ = I × (1 − c·e); acceptance then lowers the distress it produces, D = I′ × (1 − A).

Manageable — conflict is contained, the difference tolerated. When changeability is low, pushing change barely moves the residual issue, so acceptance carries the relief; raise changeability and the change strategy overtakes it — the two are complementary levers, not rivals.

The acceptance-and-change demonstration draws distress from an issue's severity, how much the couple pushes to change it, how changeable it actually is, and how far each partner accepts it. Lowering the changeability shows the behavioral strategy of pushing for change losing its grip, while raising acceptance still brings distress down — the core IBCT claim that acceptance is a separate lever, most valuable exactly where change is least available.

Each exchange, the pursuer’s demand answers the partner’s prior withdrawal and the withdrawer’s retreat answers the prior demand, both scaled by how reactively the couple responds. A repair rate — the chance a partner softens instead of escalating — damps the loop.

The loop is self-amplifying: each partner's response outweighs the last, and tension climbs toward crisis. De-escalation, not the elimination of conflict, is the first task of therapy.

The demand-withdraw demonstration iterates the cycle of Figure 1 over successive exchanges. Raising the reactivity with which each partner answers the other sends the spiral upward; introducing a repair rate — the chance that a partner softens rather than escalates — damps it. The demonstration makes visible why de-escalation, not the elimination of conflict, is the first task of therapy.

Translate a treatment’s standardized effect d into its common-language form, CL = Φ(d/√2) — the probability that a treated couple ends up better off than an untreated one — then apply a five-year maintenance rate to see how many of those gains endure.

A typical couple-therapy effect of d ≈ 0.7 means about 69% of treated couples fare better than the average untreated one; at a maintenance rate near half, roughly 38 in 100 are both helped and still improved five years on — real and durable for many, though not for all.

The outcome demonstration converts a treatment's effect size into its common-language equivalent — the probability that a treated couple ends up better off than an untreated one — and then applies a maintenance rate to show how many of those gains survive to five-year follow-up. It grounds the abstract claim that couple therapy is effective in a concrete count of couples helped, and helped durably.

Worked Example

Take the acceptance-and-change model the first demonstration draws from IBCT. Rate an issue's severity I, the couple's change effort c, the issue's changeability e, and the partners' acceptance A, each on a 0-to-1 scale. Let change effort reduce the issue only to the extent the issue is actually changeable, giving a residual issue I′ = I × (1 − c · e), and let acceptance reduce the distress that residual issue produces, giving distress D = I′ × (1 − A). The structure encodes the IBCT insight that pushing for change pays off only when change is possible, while acceptance can lower distress regardless (Roddy et al., 2016).

Consider a couple whose recurring conflict is a deep, temperamental difference — one partner needs solitude, the other closeness — so the issue is severe (I = 0.8) and barely changeable (e = 0.2). A traditional change-focused approach pushes hard on it: c = 0.9, with little acceptance, A = 0.1. The residual issue is I′ = 0.8 × (1 − 0.9 × 0.2) = 0.8 × 0.82 = 0.656, and distress is D = 0.656 × (1 − 0.1) = 0.656 × 0.9 = 0.5904. All that change effort barely dents the issue, because the issue was never negotiable, and distress stays high.

Now treat the same couple with the IBCT balance: push change less, c = 0.4, and cultivate acceptance, A = 0.7. The residual issue is actually a little higher, I′ = 0.8 × (1 − 0.4 × 0.2) = 0.8 × 0.92 = 0.736 — less change effort was spent — yet distress falls to D = 0.736 × (1 − 0.7) = 0.736 × 0.3 = 0.2208. The difference between the partners is essentially unchanged; what changed is that they stopped fighting it. The number captures the paradox at the heart of IBCT: for an unchangeable difference, distress dropped by more than half not by solving the problem but by ceasing to demand that it be solved (Christensen et al., 2004). Had the issue instead been genuinely changeable (raise e toward 0.9), the arithmetic would reverse and the change-focused strategy would win — which is why the model treats acceptance and change as complementary, not rival.

Efficacy and Durability

The evidence that couples therapy works is now substantial and quantitative. An early meta-analysis of randomized controlled trials of behavioral marital therapy established the format's efficacy against no-treatment controls, putting the first firm number on a treatment that had been supported mostly by individual trials (Shadish & Baldwin, 2005). The most recent large meta-analysis, pooling effects across outcomes, designs, timeframes, and moderators, places couple therapy among the more effective psychotherapies, with moderate-to-large effects on relationship satisfaction that are robust across the variations the analysis tested (Roddy et al., 2020).

Durability is the harder test, and here the picture is honest rather than triumphant. The five-year follow-up of the pivotal IBCT-versus-traditional trial found that a substantial share of couples maintained their gains, with roughly half still improved and a minority separated or divorced — evidence that the effects are real and lasting for many couples, without pretending they hold for all (Christensen et al., 2010). A review of the decade's treatment research reaches the same balance across the two dominant empirically supported models, IBCT and EFT, confirming both as effective while marking relapse and the couples who do not respond as the field's standing problems (Doss et al., 2022).

Discussion

Couples therapy occupies a distinctive position among psychological treatments because its object is not a person but a relationship — a two-body system whose behavior is not reducible to the psychology of either partner. That makes it a natural laboratory for social cognition: the demand-withdraw cycle is a pattern each partner reads, predicts, and responds to, and the therapy works by changing not the individuals but the way each interprets and answers the other. The reframing that locates the problem in the cycle rather than the person is, in cognitive terms, a change in attribution — from a stable trait of the partner to a dynamic each co-produces (Snyder et al., 2006).

The field's two mature turns sharpen the same point in different directions. IBCT's acceptance move is a claim about appraisal: the same unchangeable difference can be a source of chronic distress or a tolerated fact depending on how it is construed, so that a shift in meaning, not in behavior, can be curative. EFT's attachment move is a claim about emotion: that the affect driving the cycle is not incidental irritation but attachment protest, and that reaching it is what lets the pattern reorganize. Both depart from the original behavioral premise that relationships are repaired chiefly by better exchange and better skills, and both have the outcome evidence to justify the departure — which is why contemporary practice is increasingly integrative, drawing on all three models rather than adjudicating among them (Lebow & Snyder, 2022).

Current Directions

The contemporary front of couples therapy research runs in three directions. The first is dissemination and access: the recognition that effective treatments reach only a fraction of distressed couples has driven the adaptation of IBCT into web-based programs, such as the OurRelationship intervention, that deliver the core of the therapy at scale and with minimal clinician time, and the last-decade reviews treat technology-assisted delivery as a central rather than peripheral development (Doss et al., 2022). The second is precision: the most recent meta-analysis is explicitly moderator-focused, asking not merely whether couple therapy works but for whom, under which designs, and over what timeframe the effects hold, moving the field from an average effect toward a map of its boundary conditions (Roddy et al., 2020).

The third is integration and scope. A state-of-the-field review of couple therapy in the 2020s maps a practice that increasingly combines the behavioral, acceptance-based, and attachment-based models rather than choosing among them, and that is extending to diversity-responsive care and to couples presenting with individual disorders — depression, trauma, chronic illness — in which the relationship is both affected by and a resource against the condition (Lebow & Snyder, 2022). Across all three, the open questions are the durable ones the outcome research keeps returning: how to prevent relapse, how to help the couples who do not respond, and how to match model to couple rather than applying one to all.

Common Misconceptions

Couples therapy is for marriages on the brink of divorce.

It is most effective earlier, before patterns entrench. The observational research shows that what predicts dissolution is a specific style of negative interaction, which is easier to interrupt before it has hardened into the couple's default; waiting until separation is imminent leaves less to work with, not more (Gottman & Levenson, 1992).

The goal is to stop the couple from fighting.

Conflict itself does not distinguish happy from unhappy couples — both disagree. What distinguishes them is how conflict is conducted, so the therapeutic aim is to change the pattern of the exchange, not to eliminate disagreement, which is neither possible nor desirable (Snyder et al., 2006).

A good therapist changes each partner so they stop causing problems.

Some differences between partners are stable and cannot be changed, and treating them as faults to be corrected can deepen the conflict. Integrative behavioral couple therapy makes emotional acceptance a co-equal aim precisely because, for an unchangeable difference, ceasing to fight it can relieve distress more than trying to fix it (Christensen et al., 2004).

Couples therapy is unproven, soft, or merely supportive.

Meta-analyses of randomized trials place it among the more effective psychotherapies, with moderate-to-large effects on satisfaction and gains that roughly half of couples maintain at five years. It is one of the better-evidenced psychological treatments, not a lesser one (Roddy et al., 2020).

Glossary

Acceptance (emotional acceptance).

In integrative behavioral couple therapy, coming to tolerate and understand a partner's stable characteristics rather than struggling to change them; a mechanism of relief distinct from behavior change, most valuable where change is not possible.

Attachment bond.

The emotional tie between partners that emotionally focused therapy treats as the seat of relationship distress, construing conflict as a protest against a threatened loss of secure connection.

Behavior exchange.

The give-and-take of rewarding and punishing actions between partners; behavioral marital therapy construes distress as a shortage of positive exchange and treats it by increasing the rewarding behaviors each partner offers the other.

Behavioral marital therapy.

The original empirically studied model, treating distress as a shortage of rewarding behavior exchange and a deficit of communication and problem-solving skill, and remedying it by teaching both.

Common-language effect size.

A restatement of a standardized effect size as the probability that a randomly chosen treated case scores better than a randomly chosen untreated one, used here to express how many couples a therapy helps.

De-escalation.

The first task of emotionally focused therapy: interrupting the self-amplifying negative cycle so that conflict settles rather than spirals, a precondition for restructuring the bond, and the goal that matters more than eliminating disagreement.

Demand-withdraw cycle.

The self-amplifying pattern in which one partner's pursuit or demand is met by the other's withdrawal, which intensifies the demand; a core target of de-escalation in both IBCT and emotionally focused therapy.

Emotionally focused therapy (EFT).

Susan Johnson's attachment-based model, which de-escalates the negative cycle and then restructures the emotional bond through engaged encounters in which partners express and respond to attachment needs.

Group psychotherapy.

The broader MeSH class under which couples therapy is indexed, a two-person dyad being treated as the smallest therapeutic group.

Integrative behavioral couple therapy (IBCT).

Christensen and Jacobson's model, which pairs behavior change with emotional acceptance, holding that some differences cannot be negotiated away and that the struggle over them, more than the difference itself, produces distress.

Marital therapy.

The narrower MeSH descriptor and older term, directed specifically at the conflicts of a marriage; the contemporary term couples therapy generalizes it to unmarried, cohabiting, and same-sex partners.

Negative-affect reciprocity.

The tendency, heightened in distressed couples, for one partner's negative emotion to be answered with negative emotion, producing the escalating sequences that predict dissolution.

Reframing.

The therapist's move of relocating the presenting problem from a fault in one partner to the cycle the couple jointly enacts; in cognitive terms a shift of attribution from a stable trait to a dynamic both co-produce.

Relationship distress.

The clinical target of couples therapy: chronic dissatisfaction and conflict located in the interaction between partners rather than in a symptom of either individual.

Key Researchers

Donald H. Baucom. Distinguished professor at the University of North Carolina at Chapel Hill and a developer of cognitive-behavioral couple therapy, whose 1998 review established behavioral marital therapy as an empirically supported treatment. Faculty page - Wikipedia - Wikidata

Andrew Christensen. Psychologist at the University of California, Los Angeles and, with Neil Jacobson, co-developer of integrative behavioral couple therapy, which added emotional acceptance to the behavioral change model. Faculty page - Wikipedia - Wikidata

Brian D. Doss. Psychologist at the University of Miami and a leading contemporary couple-therapy researcher, who adapted IBCT into the web-based OurRelationship program and has led recent meta-analyses and decade reviews of the field. ORCID - Google Scholar - Faculty page

John M. Gottman. Emeritus professor of psychology at the University of Washington (b. 1942), whose longitudinal, observational studies of marital interaction identified the negative-affect patterns that predict dissolution. Wikipedia - Wikidata

Alan S. Gurman. Emeritus professor of psychiatry (1945-2013) at the University of Wisconsin-Madison, editor of the field's standard clinical handbook and author of its definitive historical review of couple therapy. Wikidata

Neil S. Jacobson. Psychologist (1949-1999) at the University of Washington, pioneer of behavioral marital therapy and co-developer, with Andrew Christensen, of integrative behavioral couple therapy. APS memorial

Susan M. Johnson. Clinical psychologist (1947-2024), emeritus professor at the University of Ottawa and founder of the International Centre for Excellence in Emotionally Focused Therapy, who developed emotionally focused therapy for couples on an attachment foundation. ICEEFT profile - Wikipedia

Douglas K. Snyder. Psychologist at Texas A&M University, developer of insight-oriented couple therapy and the Marital Satisfaction Inventory, and author of the 2006 Annual Review synthesis of the field. Google Scholar - Faculty page

Frequently Asked Questions

What is couples therapy?
Couples therapy is the treatment of relationship distress by working with both partners together, taking the relationship between them, the recurring patterns of interaction the two create, as the unit of care rather than a symptom in either person. The therapist locates the difficulty in the cycle the couple enacts and seeks change in how the partners respond to each other (Gurman & Fraenkel, 2002).

How is couples therapy different from individual therapy?
It attends to what happens between the partners rather than within either one. Where individual therapy works on a person's internal history, couples therapy targets the present-tense exchange, such as how a bid for closeness is met or how a complaint escalates, because that observable interaction is both the disorder and the site of its repair (Snyder et al., 2006).

What are the main types of couples therapy?
Three empirically supported models dominate: traditional behavioral couple therapy, which improves behavior exchange and communication skill; integrative behavioral couple therapy, which adds emotional acceptance to behavior change; and emotionally focused therapy, which treats distress as a disruption of the attachment bond and works to restructure it (Wiebe & Johnson, 2016).

What is integrative behavioral couple therapy?
IBCT, developed by Andrew Christensen and Neil Jacobson, keeps the behavioral aim of change but adds emotional acceptance as a co-equal goal. Its insight is that some differences between partners cannot be negotiated away, and that the polarized struggle over them, more than the difference itself, produces the distress, so acceptance becomes curative where change is not available (Christensen et al., 2004).

What is emotionally focused therapy for couples?
Emotionally focused therapy, developed principally by Susan Johnson, draws on attachment theory to treat relationship distress as a protest against a threatened loss of secure connection. It de-escalates the couple's negative cycle and then restructures the bond through emotionally engaged encounters, and its early controlled trials reported large effect sizes (Johnson et al., 1999).

Does couples therapy actually work?
Yes. Meta-analyses of randomized trials place couple therapy among the more effective psychotherapies, with moderate-to-large effects on relationship satisfaction that hold across outcomes, designs, and timeframes (Roddy et al., 2020). Its efficacy against no-treatment controls was established for behavioral marital therapy in an earlier meta-analysis (Shadish & Baldwin, 2005).

Do the gains last?
For many couples, yes, though not for all. The five-year follow-up of the pivotal IBCT trial found that a substantial share of couples maintained their gains, with roughly half still improved while a minority separated, evidence that the effects are durable for many without pretending they hold universally (Christensen et al., 2010).

How is couples therapy being made more accessible?
Because effective treatments reach only a fraction of distressed couples, researchers have adapted IBCT into web-based programs that deliver its core with minimal clinician time, and recent reviews treat technology-assisted delivery as a central development rather than a peripheral one (Doss et al., 2022).

References