Cognitive analytic therapy
(CAT)
- Therapy · Integrative
- Limited or early evidence
- Published October 19, 2026
- 17 sources
- 6 min read
In short
Related in this topic
On this page
Key facts
- Tradition
- Integrative
- Developed by
- Anthony Ryle, with theoretical development aided by Mikael Leiman
- Origin
- Developed by Ryle from research on psychotherapy process; a 2004 review traced the model’s evolution over the previous 25 years
- Typical format
- Individual and time-limited, usually 16 weekly sessions and 24 for more complex difficulties; an 8-session version has been tested for depression
- Most researched for
- Borderline personality disorder in young people; Personality disorders in adults; Depression (8-session version)
- Also known as
- Cognitive–analytic therapy
State of the evidence
Limited or early evidenceMeta-analysis finds large improvements from before to after treatment and small to moderate advantages over comparison treatments in nine trials[1], and dropout is lower than with comparison treatments[2]. In young people with borderline personality disorder, two randomized trials found no advantage for CAT over good clinical care or befriending on the main outcomes, although patients improved faster in one and attended more in the other[3][4]. An earlier review judged comparisons with other therapies largely equivocal[5].
Evidence labels summarise the research literature; they are not recommendations. How labels are assigned
Cognitive analytic therapy (CAT) is a brief, focal psychotherapy that draws on cognitive therapy, psychodynamic psychotherapy and developments in cognitive psychology. It was developed by Anthony Ryle in response to the National Health Service’s need for short treatments[6]. Ryle described it as derived from an integration of conventionally opposed models and delivered from the start within a limited time frame, usually 16 weekly sessions and 24 for more disturbed patients[7]. Its core features, in the account of the Association for Cognitive Analytic Therapy, are an early joint description of the patient’s problem patterns, the patient’s use of that description to recognize and control damaging ways of acting, and the therapist’s use of it to avoid reciprocating and reinforcing those patterns[8].
Development from general practice and research
Ryle began his career in general practice. In 1964 he became director of the University of Sussex health service, where over the next 15 years he made considerable use of Kelly’s repertory grid techniques in research. He received supervision from a psychoanalyst and practiced from an object relations and group-analytic orientation. Reservations about psychoanalysis led him toward integrating the ideas and methods of different psychotherapies[8].
To study how change happened, Ryle read the notes of completed therapies. He found the work centered on patients’ failure to revise harmful ways of thinking and acting, and grouped the failures into three patterns, which he called traps, dilemmas and snags. The association traces traps to the sequential descriptions of behavior therapy, dilemmas to the limited options that repertory grids display, and snags to psychoanalytic ideas about guilt and to family and systems approaches[8]. When these descriptions were used in the first sessions with new patients, originally as a research procedure, patients began to recognize and revise their “problem procedures,” and the therapy grew from that observation[8].
As psychotherapy consultant at St Thomas’ Hospital, Ryle adopted a policy of offering the minimum sufficient intervention. The great majority of patients, including those with personality disorders, received 12 to 16 sessions, mostly from trainees. Aided by Mikael Leiman, he developed the distinct theoretical basis of CAT and continued research, particularly on borderline personality disorder. He retired from the National Health Service in 1992 and died in September 2016, aged 89[8].
Reciprocal roles and the multiple self states model
The central concept is the reciprocal role relationship, a pattern of relating between self and other that originates in childhood. Ryle and Fawkes describe a range of such patterns that alternate with the situation, which they call contextual multiplicity. Adverse childhood experiences may restrict the range, and severe deprivation or abuse may result in a structural dissociation of self-processes[9]. In 1997 Ryle extended the theory to borderline personality disorder, proposing a structural model, the multiple self states model, with a classification of levels of developmental damage, to explain the phenomenology of the disorder[10].
Ryle set the concept against psychoanalytic accounts. In a 1994 paper he called projective identification “a particular form of reciprocal role procedure” and argued that the phenomena it describes are best understood as forms of the more general processes involved in interpersonal relating[11]. In practice, therapist and patient describe the dysfunctional patterns and the transitions between them at the start of therapy and use the descriptions throughout[9].
Reformulation letters and diagrams
The first phase, roughly sessions 1 to 5 of a 16-session therapy, is called reformulation. Therapist and patient explore early life, sometimes with a timeline. The therapist attends to the reciprocal roles the patient takes up, may use questionnaires such as the Psychotherapy File, and begins to map the roles and patterns on paper[12]. The phase ends with a reformulation letter, usually given between the fourth and eighth sessions and later in a 24-session therapy; in an 8-session therapy a letter may not be offered[13].
As the association describes it, the therapist reads the letter aloud and gives the patient a copy. It covers why the patient came, important life events, past and present relationships and wider influences such as culture and politics. It explains how the problem patterns developed, often as ways of surviving other difficulties, notes the patient’s strengths, and proposes up to three or four problem patterns to work on. It may anticipate how those patterns could appear between patient and therapist, refers to the ending of therapy and invites the patient to respond[13].
The diagram, known as a sequential diagrammatic reformulation, uses words and arrows to describe the patterns visually[13]. Beard, Marlowe and Ryle described its use in 1990 with patients with personality disorders. They suggested that careful analysis of these patients’ shifting states of mind could reduce the bewilderment the patients provoke and experience, and that diagrams tracing the shifts help both patients and clinicians[14]. In a small study of six therapist–client pairs, delivery of the letter and diagram did not consistently change measures of insight or symptoms, although clients credited the tools with insight and emotional change within a good therapeutic alliance[15] . A dismantling trial randomized 95 depressed patients in primary care to eight sessions of CAT with or without the narrative reformulation and found no difference in outcome, attendance or alliance; the authors suggested that the narrative reformulation may be redundant in treating depression[16].
The time-limited course
In the active phase, roughly sessions 6 to 12, patient and therapist use the diagram to recognize problem patterns as they occur and to develop alternatives. The work may include skills such as relaxation, self-soothing and assertiveness, a more compassionate “observing eye,” attention to how the patterns appear in the relationship with the therapist, and in some cases closer work on traumatic events[12]. The ending is anticipated from the start, and the reformulation letter refers to it[13]. The final sessions, roughly 13 to 16, address earlier losses and endings, review what has changed and acknowledge what cannot change; a follow-up appointment may be offered after about three months[12]. Toward the end the therapist offers a second letter and invites the patient to write about the experience of therapy[13].
Trials in young people with borderline personality disorder
In Melbourne, Chanen and colleagues randomized 86 outpatients aged 15 to 18 who met between two and nine criteria for borderline personality disorder to up to 24 sessions of CAT or to manualized good clinical care, both within a comprehensive service. At 24 months the groups did not differ on the chosen outcomes, although there was some evidence that the CAT group improved more rapidly; both treatments reduced externalizing psychopathology[3].
The later MOBY trial randomized 139 people aged 15 to 25 with the disorder to a dedicated early-intervention service (Helping Young People Early, or HYPE) with weekly CAT, the same service with befriending, or a general youth mental health service with befriending. All groups improved, and neither the specialist service nor CAT produced faster improvement in psychosocial functioning by 12 months. HYPE with CAT achieved better attendance and completion than befriending. The authors concluded that effective early intervention did not depend on specialist psychotherapy but did require youth-oriented case management and psychiatric care; two of them co-founded and lead the HYPE program[4].
Adult trials and reviews
In a service-based trial in adults with personality disorders, 38 participants received 24 sessions of CAT and 40 received treatment as usual. Those given CAT showed reduced symptoms and substantial benefits compared with the comparison group, which showed signs of deterioration; the authors called for more elaborate and controlled evaluations[17].
A 2014 review found 25 outcome studies, five of them randomized. Most were small, practice-based studies of complex and severe presentations, 44% concerned personality disorder, and where CAT had been compared with other therapies the results were largely equivocal[5]. A 2021 meta-analysis of 25 studies found large improvements from before to after treatment in global functioning (effect size 0.86) and depression (1.05), moderate to large improvement in interpersonal problems (0.74) and small to moderate advantages over comparison treatments in nine trials (0.36 to 0.53); average dropout was 16%[1]. A meta-analysis of 34 studies put treatment refusal at 15% and dropout at 19%, with lower dropout than comparison treatments in the same studies[2]. Kellett, an author of the dismantling trial, is also an author of all three reviews[16][5][1][2].
Sources17
Numbered in order of first citation.
- 1.Hallam, C., Simmonds-Buckley, M., Kellett, S., Greenhill, B., & Jones, A. (2021). The acceptability, effectiveness, and durability of cognitive analytic therapy: Systematic review and meta‐analysis. Psychology and Psychotherapy: Theory, Research and Practice, 94(S1), 8–35. https://doi.org/10.1111/papt.12286
- 2.Simmonds-Buckley, M., Osivwemu, E. O., Kellett, S., & Taylor, C. (2022). The acceptability of cognitive analytic therapy (CAT): Meta-analysis and benchmarking of treatment refusal and treatment dropout rates. Clinical Psychology Review, 96, 102187. https://doi.org/10.1016/j.cpr.2022.102187
- 3.Chanen, A. M., Jackson, H. J., McCutcheon, L. K., Jovev, M., Dudgeon, P., Yuen, H. P., Germano, D., Nistico, H., McDougall, E., Weinstein, C., Clarkson, V., & McGorry, P. D. (2008). Early intervention for adolescents with borderline personality disorder using cognitive analytic therapy: randomised controlled trial. British Journal of Psychiatry, 193(6), 477–484. https://doi.org/10.1192/bjp.bp.107.048934
- 4.Chanen, A. M., Betts, J. K., Jackson, H., Cotton, S. M., Gleeson, J., Davey, C. G., Thompson, K., Perera, S., Rayner, V., Andrewes, H., & McCutcheon, L. (2022). Effect of 3 forms of early intervention for young people with borderline personality disorder: The MOBY randomized clinical trial. JAMA Psychiatry, 79(2), 109–119. https://doi.org/10.1001/jamapsychiatry.2021.3637
- 5.Calvert, R., & Kellett, S. (2014). Cognitive analytic therapy: A review of the outcome evidence base for treatment. Psychology and Psychotherapy: Theory, Research and Practice, 87(3), 253–277. https://doi.org/10.1111/papt.12020
- 6.Denman, C. (2001). Cognitive–analytic therapy. Advances in Psychiatric Treatment, 7(4), 243–252. https://doi.org/10.1192/apt.7.4.243
- 7.Ryle, A. (2004). The contribution of cognitive analytic therapy to the treatment of borderline personality disorder. Journal of Personality Disorders, 18(1), 3–35. https://doi.org/10.1521/pedi.18.1.3.32773
- 8.Association for Cognitive Analytic Therapy (n.d.). How CAT developed. Association for Cognitive Analytic Therapy. www.acat.org.uk/cat-explained/how-cat-developed
- 9.Ryle, A., & Fawkes, L. (2007). Multiplicity of selves and others: Cognitive analytic therapy. Journal of Clinical Psychology, 63(2), 165–174. https://doi.org/10.1002/jclp.20339
- 10.Ryle, A. (1997). The structure and development of borderline personality disorder: A proposed model. British Journal of Psychiatry, 170(1), 82–87. https://doi.org/10.1192/bjp.170.1.82
- 11.Ryle, A. (1994). Projective identification: A particular form of reciprocal role procedure. British Journal of Medical Psychology, 67(2), 107–113. https://doi.org/10.1111/j.2044-8341.1994.tb01776.x
- 12.Jenaway, A. (n.d.). What CAT sessions are like. Association for Cognitive Analytic Therapy. www.acat.org.uk/cat-explained/what-cat-sessions-are-like
- 13.Association for Cognitive Analytic Therapy (n.d.). Letters in CAT: The first letter. Association for Cognitive Analytic Therapy. www.acat.org.uk/cat-explained/what-cat-sessions-are-like/letters-in-cat-the-first-letter
- 14.Beard, H., Marlowe, M., & Ryle, A. (1990). The management and treatment of personality-disordered patients: The use of sequential diagrammatic reformulation. British Journal of Psychiatry, 156(4), 541–545. https://doi.org/10.1192/bjp.156.4.541
- 15.Tyrer, R., & Masterson, C. (2019). Clients' experience of change: An exploration of the influence of reformulation tools in cognitive analytic therapy. Clinical Psychology & Psychotherapy, 26(2), 167–174. https://doi.org/10.1002/cpp.2339
- 16.Kellett, S., Stockton, C., Marshall, H., Hall, J., Jennings, C., & Delgadillo, J. (2018). Efficacy of narrative reformulation during cognitive analytic therapy for depression: Randomized dismantling trial. Journal of Affective Disorders, 239, 37–47. https://doi.org/10.1016/j.jad.2018.05.070
- 17.Clarke, S., Thomas, P., & James, K. (2013). Cognitive analytic therapy for personality disorder: randomised controlled trial. British Journal of Psychiatry, 202(2), 129–134. https://doi.org/10.1192/bjp.bp.112.108670
Further reading
- Ryle, A. (2004). The contribution of cognitive analytic therapy to the treatment of borderline personality disorder. Journal of Personality Disorders, 18(1), 3–35. https://doi.org/10.1521/pedi.18.1.3.32773
- Denman, C. (2001). Cognitive–analytic therapy. Advances in Psychiatric Treatment, 7(4), 243–252. https://doi.org/10.1192/apt.7.4.243
- Hallam, C., Simmonds-Buckley, M., Kellett, S., Greenhill, B., & Jones, A. (2021). The acceptability, effectiveness, and durability of cognitive analytic therapy: Systematic review and meta‐analysis. Psychology and Psychotherapy: Theory, Research and Practice, 94(S1), 8–35. https://doi.org/10.1111/papt.12286
Related entries 6
- Object relations theoryConcept
- Cognitive behavioral therapy(CBT)Therapy
- Psychodynamic therapy(PDT)Therapy
- Projective identificationConcept
- Therapeutic allianceConcept
- Case formulationMethod
Cite this entry
APA 7
Psych Dictionary. (2026, October 19). Cognitive analytic therapy (CAT). In Psych Dictionary. https://www.psychdictionary.com/therapies/cognitive-analytic-therapy
MLA 9
“Cognitive Analytic Therapy (CAT).” Psych Dictionary, 19 Oct. 2026, www.psychdictionary.com/therapies/cognitive-analytic-therapy.