Does Couples Therapy Work? An Honest Look at the Evidence

Does Couples Therapy Work? An Honest Look at the Evidence

This is the kind of question that gets typed late at night, sometimes with the word "actually" in it. It deserves a real answer rather than a booking pitch — partly because the research is genuinely encouraging, and partly because the honest caveats explain why some couples try therapy and walk away saying it did nothing.

This article covers what the studies actually found, where the famous Gottman research fits and where its limits are, why couples therapy fails when it fails, and what makes it more likely to help.

Does Couples Therapy Actually Work?

The short version: in controlled trials, couples who received therapy were better off, on average, than similar couples who received none (Shadish and Baldwin, 2005).

The longer version has three caveats attached, and they matter more than the headline. First, "on average" is doing real work in that sentence — a meaningful share of couples in the trials did not improve. Second, many trials measure relationship satisfaction over months, not whether a marriage is thriving a decade later. Third, the couples in trials showed up and completed treatment; in ordinary life, some couples arrive very late and quit early, which changes the picture. Each of these gets its own section below.

What Does the Research Actually Say?

The broadest evidence comes from meta-analyses, which pool many trials. William Shadish and Scott Baldwin's 2005 meta-analysis of 30 randomized trials in the Journal of Consulting and Clinical Psychology found behavioural couple therapy significantly more effective than no treatment, while warning that publication bias may have inflated earlier estimates (Shadish and Baldwin, 2005). Their 2003 review of 20 meta-analyses likewise supported the efficacy of marital and family therapy for distressed couples, with effects slightly reduced at follow-up but still significant (Shadish and Baldwin, 2003).

On durability, a particularly informative study is Andrew Christensen and colleagues' randomized trial (2004, Journal of Consulting and Clinical Psychology), which followed 134 chronically and seriously distressed couples through two forms of behavioural couple therapy. At the five-year follow-up (Christensen, Atkins, Baucom and Yi, 2010), roughly half of the couples in each treatment showed clinically significant improvement over where they started. That number is worth sitting with, because it is both encouraging and sobering: these were couples in serious, long-standing distress, and about half were doing measurably better five years on — and about half were not.

Emotionally Focused Therapy, developed by Sue Johnson and Les Greenberg, has its own trial base. A frequently cited meta-analysis (Johnson, Hunsley, Greenberg and Schindler, 1999, Clinical Psychology: Science and Practice) reviewed the early randomized trials and presented meta-analytic data on improvement in couple adjustment with EFT. The honest caveat, which applies well beyond EFT: many of those trials were small and run by researchers connected to the model itself, and the field calls that researcher allegiance — an overview of 30 meta-analyses found a moderate, robust association between researcher allegiance and reported treatment effects (Munder and colleagues, 2013).

Put together, a fair summary reads like this: couples therapy is neither a coin flip nor a cure-all. The evidence supports it as a genuine treatment with real effects — and it also says that nobody can honestly tell you in advance which side of the average you will land on. What you can influence is covered in the sections that follow.

Is the Gottman Method Evidence-Based?

The Gottman name comes up in almost every search about couples therapy, so it is worth being precise about what that research is and is not.

What John Gottman and Robert Levenson actually did, beginning in the 1970s, was observational science: they recorded couples in conflict conversations, tracked physiology, and followed the same couples for years (for example, Gottman and Levenson, 1992). Out of that came the field's most recognizable descriptive findings — including the Four Horsemen (criticism, contempt, defensiveness, stonewalling) as patterns that distinguished couples who later separated. The Gottman Institute publishes summaries of this body of work.

Two honest qualifications. First, that research is correlational: it describes what distressed couples do, which is not the same as proving that coaching couples out of those patterns saves relationships — the interventions built on the observations have their own, smaller evidence base. Second, the famous claims about predicting divorce with striking accuracy came from models fitted to data after the outcomes were known; Richard Heyman and Amy Slep (2001, Journal of Marriage and Family) showed that the accuracy of this kind of prediction drops sharply when it is cross-validated. None of that makes the observational work less valuable — it is a detailed record of how real couples interact — but "correlational research with interventions built on top" is the accurate label, and a practice that uses the method should be able to say so plainly.

At TEO Counselling, Mohamad Shabib has completed Gottman Method training at Levels 1 and 2, and you can read how it is used in practice on the Gottman Method page.

Why Does Couples Therapy Fail for Some Couples?

Timing, first and most of all. The Gottman Institute has long cited John Gottman's figure that couples wait an average of six years of being unhappy before getting help (The Gottman Institute). The figure is attributed to Gottman himself rather than to an independent study, so treat it as his estimate — but the pattern it describes is a familiar one in couples work. By the time some couples arrive, contempt has had years to calcify, and sometimes one partner has already privately left. Therapy that might have shifted a two-year-old pattern is asked to reverse a ten-year-old one.

A mixed agenda in the room. Some couples arrive with one partner working on the relationship and the other quietly working on the exit. Standard couples therapy assumes two people pulling in roughly the same direction; when that assumption is false, sessions stall in a way that looks like therapy failing. A good therapist names this early, because a couple deciding whether to work on the marriage needs a different conversation than a couple deciding how.

Quitting early. Trials report outcomes for couples who completed treatment; in everyday practice, plenty of couples stop after two or three sessions — sometimes because sessions started to feel like the same argument with an audience. That experience is worth reporting to the therapist rather than acting on silently, because restructuring how sessions run is a normal, expected adjustment, not a sign the whole project is doomed.

Something else running alongside. Untreated addiction, or a mental health condition currently overwhelming one partner, can quietly cap what joint sessions can achieve until it is addressed. Sequencing matters, and a consultation call is a reasonable place to ask what order things should happen in.

Therapist fit. Couples work is a distinct skill, and some generalist therapists see couples without couples-specific training. Asking directly — what couples training do you have, and is it training or certification? — is a fair question that a good therapist will welcome.

How Long Does Couples Therapy Take to Work?

No responsible therapist will give you a fixed number in advance, because it depends on how long the pattern has run and what it is protecting. What you can reasonably expect from structured couples work: an assessment phase over the first few sessions, a shared plan agreed early rather than open-ended talking, something concrete to practise between sessions, and honest check-ins about whether it is helping. If several sessions in nothing at all is shifting, saying exactly that in the room is part of the work — not a rude interruption of it.

Does Couples Therapy Work After an Affair?

Affairs are one reason couples book, and recovery is possible — but it is a structured rebuild, not a single clearing-of-the-air conversation. The work can run through phases: dealing with the immediate impact, understanding honestly how the relationship got here, and then deciding — both people, freely — whether to rebuild trust and on what terms. No honest therapist promises a particular ending. What therapy offers is a structure in which either ending can be reached deliberately instead of by attrition. Affairs that never became physical raise their own questions, about what counts as a breach and whether trust can come back, and our separate guide to recovering after an emotional affair takes those up.

Does Online Couples Therapy Work?

A 2022 brief by the Ontario COVID-19 Science Advisory Table, summarized by CMHA Ontario, found virtual mental health care efficacious for depression, anxiety and post-traumatic stress disorder in comparison with in-person care; it was about mental health care for adults generally, not couples therapy in particular. For couples, online sessions remove a common practical obstacle — finding one protected hour that two working adults can both reach — and let partners join from separate locations when travel would otherwise mean cancelling. TEO Counselling's couples counselling is virtual by design, with evening and weekend appointments available.

When Couples Therapy Is Not the Right Tool

Where there is abuse — violence, coercive control, intimidation, or fear of a partner's reaction — joint sessions can be unsafe, because speaking honestly in front of the partner can carry consequences after the call ends. In those situations, individual support is the safer starting point, and a therapist should help arrange exactly that rather than pressing on with joint work.

If you are in crisis, call or text 9-8-8, the 9-8-8 Suicide Crisis Helpline, available across Canada at no cost. In British Columbia you can also call 1-800-SUICIDE (1-800-784-2433). If you are in immediate danger, call 911.

What Makes Couples Therapy More Likely to Help?

• →Go earlier than feels necessary. The six-year delay is one of the more fixable factors on this page. A pattern gets easier to change before it becomes the relationship's architecture.
• →Both partners on the consultation call. A reluctant partner asking their own skeptical questions beats a reluctant partner being reported on.
• →Ask about couples-specific training. Approach, training level, and whether it is training or certification — a therapist worth booking answers all three without flinching.
• →Treat the between-session practice as the therapy. The hour is where patterns get mapped; the week is where they change.
• →Say the doubts in the room. "This feels like the same fight with an audience" is useful clinical information, and raising it can change how sessions run.

An Honest First Step

You do not need to be sure therapy will work to have one conversation about whether it might. TEO Counselling offers a free 15-minute consultation — a chance for both of you to ask about approach, training, and fees, with no pressure to commit. We work with clients across British Columbia, Ontario, Alberta, Saskatchewan, Manitoba, and Newfoundland and Labrador. Book a free consultation through the Jane booking platform, or text (519) 760-5211. Evening and weekend appointments are available.

You don't have to carry it alone.