Therapeutic alliance
- Concept · Psychotherapy research
- Moderate evidence
- Published October 19, 2026
- 12 sources
- 5 min read
In short
Related in this topic
On this page
Key facts
- Field
- Psychotherapy research
- Originators
- Psychoanalytic writers including Elizabeth Zetzel and Ralph Greenson; generalized to all therapies by Edward S. Bordin
- First described
- 1950s–1979
- Also known as
- Working alliance; Helping alliance
- Topics
- How psychotherapy works
State of the evidence
Moderate evidenceThe association with outcome is robust: a meta-analysis of 295 studies with more than 30,000 patients found r = .28, consistent across treatments and rater perspectives[1], and the association survived adjustment for patient characteristics and other treatment processes[2]. Whether the alliance causes improvement is less settled. Session-by-session studies find effects in both directions[3], a study of cognitive therapy for depression found that the alliance did not predict later symptom change[4], and the association is carried mainly by differences between therapists rather than between patients[5].
Evidence labels summarise the research literature; they are not recommendations. How labels are assigned
The therapeutic alliance is the collaborative part of the relationship between a client and a therapist. In Edward Bordin’s definition it has three elements: agreement on the goals of treatment, agreement on the tasks through which those goals will be pursued, and a personal bond made up of reciprocal positive feelings[6][7]. Bordin intended the definition to apply to any form of therapy, and it has been called a pan-theoretical concept for that reason[7].
The alliance is narrower than the therapeutic relationship as a whole. It differs from transference, which in psychoanalytic theory is the client’s reaction to the therapist as a figure from the past: the analysts who introduced the alliance defined it as the realistic, non-transferential part of the relationship[7]. It also differs from the therapist-offered conditions of Carl Rogers’s person-centered therapy, such as empathy and unconditional positive regard, which describe what the therapist provides rather than what client and therapist build together[7]. “Working alliance”, “helping alliance” and “therapeutic alliance” are used for the same construct[1].
Psychoanalytic precursors
Horvath and Luborsky traced the idea to Freud’s writing on transference. Freud first regarded transference as purely negative, but later allowed that a beneficial attachment to the analyst could develop that was not simply a projection[7]. Elizabeth Zetzel, in 1956, described the therapeutic alliance as a non-neurotic, non-transferential component of the relationship that allows the patient to follow the analyst and use the analyst’s interpretations. Ralph Greenson, in 1965, defined the working alliance as a reality-based collaboration between patient and analyst[7]. These writers separated the alliance from transference, and that distinction later spread beyond psychoanalysis[7].
Lester Luborsky developed the concept further in 1976. He described two forms of alliance: an early one, based on the patient’s experience of the therapist as supportive, and a later one, in which patient and therapist share responsibility for working toward the goals of therapy[7].
Bordin’s goals, tasks and bond
Bordin’s 1979 paper moved the concept out of psychoanalysis by describing the alliance in terms that any therapy could use[6][7]. In his account the alliance is strongest when client and therapist share beliefs about the goals of treatment and regard its methods as relevant and effective, and when each accepts and carries out their own tasks. Agreement on goals and tasks depends in turn on a bond: the client must believe in the therapist’s ability to help, and the therapist in the client’s resources[7]. Bordin also proposed that the alliance affects outcome not because it heals in its own right but because it enables the client to accept, follow and believe in the treatment[7], a position that anticipates the later debate about causation.
The Working Alliance Inventory and other measures
The Working Alliance Inventory, developed by Adam Horvath and Leslie Greenberg, measures Bordin’s three dimensions with 36 items rated on seven-point scales, in versions for the client, the therapist and an observer; Tracey and Kokotovic developed a 12-item short form[8][7]. When confirmatory factor analysis in two samples (231 and 235 participants) failed to support the hypothesized structure of either form, Hatcher and Gillaspy built a revised 12-item version that separated goal, task and bond more clearly and correlated well with other alliance measures[9].
Other instruments include the California Psychotherapy Alliance Scales and the Penn scales. These scales differ in length and in the number of dimensions they measure, but they correlate well with one another, which suggests that they assess the same underlying process[7]. The alliance also changes over a course of therapy. Studies have described stable patterns, linear growth and a U-shaped pattern in which the alliance dips and then rises[7].
Association with treatment outcome
The alliance is one of the most investigated variables in psychotherapy research. A 2018 meta-analysis of 295 independent studies, covering more than 30,000 patients in studies published from 1978 to 2017, found an overall correlation between alliance and outcome of r = .278 (equivalent to d = 0.58) in face-to-face therapy, and almost the same correlation in internet-based therapy (r = .275, 23 studies). The association held across rater perspectives, alliance and outcome measures, treatment approaches, patient characteristics and countries, although effect sizes varied and 2% were negative[1].
Who contributes to the alliance matters. In 331 patients seen by 80 therapists, differences between therapists in the alliances they formed predicted outcome, whereas differences between the patients of the same therapist did not[5].
Cause, consequence or both
A correlation between alliance and outcome is compatible with three readings: that the alliance is therapeutic in itself, that it is a by-product of improvement, or that it is a common factor that allows the specific ingredients of a treatment to work. For many years correlational studies were the only evidence, and supporters of each reading cited the same correlations[10].
Early process research raised the possibility that the alliance follows improvement rather than leading it. In cognitive therapy for depression, a replication of an earlier process study found that concrete, theory-specified therapist actions early in treatment predicted later symptom change, whereas the alliance did not; prior symptom change predicted the later alliance at a trend level[4]. Later studies used methods that separate change within a patient from differences between patients. In 646 primary care patients, the alliance in one session predicted symptoms at the next, and earlier symptom change also predicted the alliance, a reciprocal pattern that the authors read as showing the alliance is not only a by-product of improvement[3]. A 2020 meta-analysis of 60 samples with 6,061 participants found that alliance-outcome correlations remained moderate (r = .23 to .31) after adjustment for intake characteristics and for simultaneous processes such as therapist adherence and competence[2].
Zilcha-Mano proposed a way to reconcile the findings. A trait-like component, the patient’s general capacity to form satisfying relationships, enables treatment to work, whereas a state-like component, change in that capacity developed through the relationship with the therapist, is what would make the alliance therapeutic in itself[10].
Ruptures and their repair
A rupture is a deterioration in the alliance, shown by disagreement about goals, a lack of collaboration on tasks or a strain in the bond[11]. Jeremy Safran, J. Christopher Muran and colleagues describe ruptures as inevitable episodes of tension or breakdown in the collaborative relationship, and reviewed evidence that their repair is related to outcome[12]. In a 2018 meta-analysis of 11 studies with 1,314 patients, rupture repair was moderately associated with good outcome (r = .29). A second meta-analysis, of six studies with 276 trainees or supervisees, found no significant effect of training therapists in rupture resolution on their patients’ outcomes (r = .11); the association was stronger when training was closer to cognitive behavioral therapy, when treatment was brief and when fewer patients had personality disorders[11].
The rupture findings share the limits of the wider literature. They are mostly correlational, they come from a small number of studies, and the practical question they raise, whether training therapists to repair ruptures improves their patients’ outcomes, has not yet been answered in favor of training[11].
Sources12
Numbered in order of first citation.
- 1.Flückiger, C., Del Re, A. C., Wampold, B. E., & Horvath, A. O. (2018). The alliance in adult psychotherapy: A meta-analytic synthesis. Psychotherapy, 55(4), 316–340. https://doi.org/10.1037/pst0000172
- 2.Flückiger, C., Del Re, A. C., Wlodasch, D., Horvath, A. O., Solomonov, N., & Wampold, B. E. (2020). Assessing the alliance–outcome association adjusted for patient characteristics and treatment processes: A meta-analytic summary of direct comparisons. Journal of Counseling Psychology, 67(6), 706–711. https://doi.org/10.1037/cou0000424
- 3.Falkenström, F., Granström, F., & Holmqvist, R. (2013). Therapeutic alliance predicts symptomatic improvement session by session. Journal of Counseling Psychology, 60(3), 317–328. https://doi.org/10.1037/a0032258
- 4.Feeley, M., DeRubeis, R. J., & Gelfand, L. A. (1999). The temporal relation of adherence and alliance to symptom change in cognitive therapy for depression. Journal of Consulting and Clinical Psychology, 67(4), 578–582. https://doi.org/10.1037/0022-006X.67.4.578
- 5.Baldwin, S. A., Wampold, B. E., & Imel, Z. E. (2007). Untangling the alliance-outcome correlation: Exploring the relative importance of therapist and patient variability in the alliance. Journal of Consulting and Clinical Psychology, 75(6), 842–852. https://doi.org/10.1037/0022-006X.75.6.842
- 6.Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research & Practice, 16(3), 252–260. https://doi.org/10.1037/h0085885
- 7.Ardito, R. B., & Rabellino, D. (2011). Therapeutic alliance and outcome of psychotherapy: Historical excursus, measurements, and prospects for research. Frontiers in Psychology, 2, 270. https://doi.org/10.3389/fpsyg.2011.00270
- 8.Horvath, A. O., & Greenberg, L. S. (1989). Development and validation of the Working Alliance Inventory. Journal of Counseling Psychology, 36(2), 223–233. https://doi.org/10.1037/0022-0167.36.2.223
- 9.Hatcher, R. L., & Gillaspy, J. A. (2006). Development and validation of a revised short version of the working alliance inventory. Psychotherapy Research, 16(1), 12–25. https://doi.org/10.1080/10503300500352500
- 10.Zilcha-Mano, S. (2017). Is the alliance really therapeutic? Revisiting this question in light of recent methodological advances. American Psychologist, 72(4), 311–325. https://doi.org/10.1037/a0040435
- 11.Eubanks, C. F., Muran, J. C., & Safran, J. D. (2018). Alliance rupture repair: A meta-analysis. Psychotherapy, 55(4), 508–519. https://doi.org/10.1037/pst0000185
- 12.Safran, J. D., Muran, J. C., & Eubanks-Carter, C. (2011). Repairing alliance ruptures. Psychotherapy, 48(1), 80–87. https://doi.org/10.1037/a0022140
Further reading
- Flückiger, C., Del Re, A. C., Wampold, B. E., & Horvath, A. O. (2018). The alliance in adult psychotherapy: A meta-analytic synthesis. Psychotherapy, 55(4), 316–340. https://doi.org/10.1037/pst0000172
- Zilcha-Mano, S. (2017). Is the alliance really therapeutic? Revisiting this question in light of recent methodological advances. American Psychologist, 72(4), 311–325. https://doi.org/10.1037/a0040435
Related entries 20
- TransferenceConcept
- CountertransferenceConcept
- Unconditional positive regardConcept
- Person-centered therapyTherapy
- Psychodynamic therapy(PDT)Therapy
- Interpersonal psychotherapy(IPT)Therapy
- Common factorsConcept
- Alliance rupture and repairConcept
- How psychotherapy worksTopic
- EmpathyConcept
- Empathy vs sympathyComparison
- Corrective emotional experienceConcept
- Object relations theoryConcept
- Placebo effect (psychology)Concept
- Empty chair techniqueMethod
- Reflective listeningMethod
- Socratic questioningMethod
- Cognitive analytic therapy(CAT)Therapy
- Gestalt therapyTherapy
- Group therapyTherapy
Cite this entry
APA 7
Psych Dictionary. (2026, October 19). Therapeutic alliance. In Psych Dictionary. https://www.psychdictionary.com/concepts/therapeutic-alliance
MLA 9
“Therapeutic Alliance.” Psych Dictionary, 19 Oct. 2026, www.psychdictionary.com/concepts/therapeutic-alliance.