Dialectical behaviour therapy (DBT) is a structured form of psychotherapy that combines behavioural methods for changing harmful patterns with strategies for accepting present experience. Developed by the American psychologist Marsha M. Linehan and colleagues, it was initially designed for people with persistent suicidal behaviour, particularly those meeting criteria for borderline personality disorder. Its central therapeutic tension is that a person’s distress and responses can be understandable while changes in behaviour remain necessary.
DBT belongs to the broader family of cognitive behavioural therapies, but comprehensive DBT is more than a set of coping techniques. It ordinarily combines individual therapy, group skills training, between-session coaching and a therapist consultation team. Treatment follows an explicit hierarchy that gives priority to life-threatening behaviour, then problems interfering with treatment, and then difficulties affecting everyday life.
The strongest established evidence concerns recurrent self-harm and suicidal behaviour in people with borderline personality disorder, with additional evidence for adolescents at high risk of self-harm. Adaptations have been developed for other clinical populations, although the strength of evidence varies substantially. Distinguishing comprehensive DBT from skills-only programmes and other DBT-informed services is essential when interpreting research or describing clinical provision.
Historical development and intellectual foundations
DBT emerged from Linehan’s work during the late 1970s and 1980s with people experiencing repeated suicidal crises. Conventional behavioural approaches offered methods for analysing and changing behaviour, but an exclusive emphasis on change could be experienced as criticism or as a failure to recognise severe suffering. Conversely, acceptance without a clear programme of behavioural change did not adequately address dangerous actions. DBT developed as an attempt to hold these therapeutic requirements together rather than alternate between them without a coherent framework.
Its behavioural foundations include learning theory, functional analysis, problem-solving, exposure and the systematic use of reinforcement. Its acceptance-based elements include validation and practices influenced by Zen and other contemplative traditions. Although some of these practices have religious historical origins, clinical DBT presents them as skills for observing experience and responding effectively, not as requirements to adopt religious beliefs. The approach’s use of mindfulness is integrated with behavioural treatment rather than offered as a separate spiritual discipline.
The word dialectical refers to the treatment’s emphasis on interdependence, ongoing change and the possibility of integrating apparently opposing positions. Acceptance and change are its best-known polarity, but others arise throughout treatment: independence and help-seeking, emotional vulnerability and personal responsibility, or consistency and flexibility. A dialectical response does not simply split the difference between two views. It seeks a more adequate understanding that preserves valid elements of each and changes an unproductive opposition.
A landmark randomised controlled trial, published by Linehan and colleagues in 1991, compared DBT with treatment as usual among 44 women with borderline personality disorder and recurrent parasuicidal behaviour. The terminology then used grouped together behaviours that later research more carefully distinguished by suicidal intent. The study reported advantages for DBT on several outcomes, including the frequency and medical severity of parasuicidal behaviour, treatment retention and psychiatric inpatient use. Its small sample and particular clinical population nevertheless limited the breadth of conclusions.
Linehan’s 1993 treatment manual and accompanying skills-training manual helped establish a reproducible intervention. Subsequent research tested DBT against stronger comparison treatments, examined its components and extended it beyond its original adult outpatient population. A substantially revised skills-training manual appeared in 2015. This history matters because “DBT” now names both a specific comprehensive treatment and a wider family of adaptations whose content, duration and supporting evidence are not identical.
The biosocial theory and proposed mechanisms
DBT’s principal explanatory framework is the biosocial theory of emotion dysregulation. In its original formulation, persistent difficulties arise through transactions between biological emotional vulnerability and an invalidating environment. Emotional vulnerability includes heightened sensitivity to emotional cues, intense emotional responses and a relatively slow return to baseline. These features are dimensional characteristics rather than a diagnostic test or a claim that every person receiving DBT has the same temperament.
An invalidating environment repeatedly dismisses, misinterprets or responds inconsistently to a person’s private experience. Examples include treating expressions of distress as inherently unreasonable, rewarding emotional escalation after ignoring less intense communication, or presenting difficult problems as if they should be easy to solve. Invalidation can occur without deliberate cruelty. The theory concerns recurring patterns of interaction and does not establish that a particular family caused a person’s disorder. Social circumstances, trauma, neurodevelopment and relationships outside the family may also shape emotional learning.
The model is transactional: the individual and environment influence one another over time. A highly reactive child may elicit confused or inconsistent responses; those responses may further impair the child’s ability to identify feelings, tolerate arousal and communicate needs. Repeated interactions can contribute to cycles in which increasingly intense behaviour becomes a more reliable way to obtain recognition or escape overwhelming situations. These are hypotheses about development and maintenance, not reasons to regard distress as calculated or insincere.
A central behavioural mechanism is negative reinforcement. A harmful behaviour may become more likely if it rapidly removes an aversive state, such as unbearable emotional arousal. The immediate relief can strengthen the behaviour even when its later consequences are damaging. Other consequences, including changes in interpersonal demands or access to care, can also influence behaviour. DBT therefore examines what precedes and follows a particular action rather than assigning a single motive to everyone who engages in it.
Treatment is intended to interrupt these cycles through several interacting processes: increased skills, more accurate emotional identification, reduced avoidance, different environmental consequences and a working therapeutic relationship. Studies have associated increased skills use with improved outcomes, but such associations do not by themselves establish a complete causal mechanism. Small neuroimaging studies have investigated changes in emotional processing, yet no biomarker is established for selecting patients for DBT or determining that it has worked. The biosocial account remains a clinical theory supported in parts by broader research, rather than a fully verified explanation of every presentation.
Comprehensive treatment structure and priorities
Standard comprehensive outpatient DBT usually includes four coordinated treatment modes. Individual psychotherapy addresses the person’s behavioural priorities and applies skills to specific events. Group skills training teaches a structured curriculum. Between-session coaching supports the use of skills in everyday situations. A therapist consultation team supports clinicians in maintaining the treatment model and working effectively with complex clinical problems. An initial course commonly lasts about a year, although programme duration and the organisation of skills teaching vary.
These modes serve related but distinct functions. Individual sessions enhance motivation and analyse obstacles to change; skills training builds capabilities; coaching helps those capabilities generalise beyond the therapy room; and consultation supports therapist competence and adherence. The treatment also addresses environmental conditions that maintain problems. A skills group alone may provide useful learning, but it does not automatically reproduce the functions or evidence base of the full programme.
DBT organises treatment through a hierarchy of behavioural targets. Life-threatening behaviours, including suicide attempts and relevant self-injury, come first. Next are therapy-interfering behaviours, which can involve either patient or therapist: repeated absence, failure to address agreed priorities, or clinician responses that undermine collaboration. Quality-of-life-interfering behaviours follow, including substance-related problems, severe relationship instability or other patterns that obstruct daily functioning. Skills acquisition accompanies this work. The hierarchy determines what receives attention first; it does not imply that lower-ranked suffering is unimportant.
Treatment stages provide a broader framework than the session-by-session hierarchy. Stage 1 addresses severe behavioural dyscontrol and the establishment of basic stability. Later stages concern persistent emotional suffering, ordinary problems of living and, in some formulations, a greater sense of connection or fulfilment. These stages are not rigid calendar periods, and patients do not necessarily complete them in a uniform sequence. Much of the clinical-trial evidence concerns Stage 1 rather than the entire proposed trajectory.
Before active treatment, a pretreatment phase usually clarifies goals, expectations and commitment. During treatment, diary cards record selected behaviours, emotions, urges and skills use, providing material for focused discussion rather than relying entirely on retrospective impressions. Between-session coaching has agreed purposes and boundaries; it is not synonymous with unlimited therapist availability or an emergency service. Comprehensive programmes therefore require defined arrangements for crises and coordination with other care. DBT is a psychotherapy, not a medication protocol, although participants may receive medication or other treatment for co-occurring conditions.
The four skills-training domains
DBT skills training is conventionally divided into mindfulness, distress tolerance, emotion regulation and interpersonal effectiveness. The domains overlap, but they address different functional problems: noticing what is happening, surviving crises without worsening them, influencing emotional patterns and navigating relationships. Teaching generally involves explanation, demonstration, practice and review of experiences outside sessions. Learning the vocabulary is not treated as equivalent to being able to use a skill during intense distress.
Mindfulness skills concern the deliberate observation of present experience. The curriculum distinguishes observing, describing and participating from the manner in which these activities occur: non-judgmentally, one-mindfully and effectively. “Wise mind” is a teaching concept for integrating emotional information with reasoned understanding. It is not a discrete neurological state. The aim is not to eliminate thought or emotion, but to reduce automatic reactions and improve awareness of the choices available in a particular situation.
Distress-tolerance skills address periods in which immediate change is impossible or attempts to force change would make matters worse. Crisis-survival material includes attentional redirection, sensory soothing and short-term methods for managing intense arousal. Acceptance-oriented material includes “radical acceptance,” meaning recognition of reality as it currently exists. Acceptance in this sense is distinguished from approval, resignation or agreement that an injustice is acceptable. These skills are intended to create room for effective action, not to substitute indefinitely for addressing solvable problems.
Emotion-regulation training develops the ability to identify emotions, understand their functions and distinguish prompting events from interpretations. “Checking the facts” examines whether an emotional response fits the available information. “Opposite action” concerns changing action tendencies when an emotion or its intensity does not fit the facts, or when acting on the emotion would be ineffective. Other material addresses vulnerability associated with disrupted routines and the development of positive experiences. Behavioural change is selected according to context; DBT does not assume that all strong emotions are mistaken.
Interpersonal-effectiveness skills distinguish objectives, relationships and self-respect. The mnemonic DEAR MAN structures requests and refusals through describing, expressing, asserting and reinforcing, alongside maintaining attention, appearing confident and negotiating. GIVE highlights behaviours intended to preserve relationships, while FAST concerns self-respect and fidelity to values. These are teaching aids rather than universal scripts. Their application depends on factors such as power differences, personal safety, cultural expectations and whether a relationship is reciprocal. Skills training therefore includes assessing the situation, not merely repeating a formula.
Behavioural analysis, validation and therapeutic strategies
One of DBT’s central procedures is behavioural chain analysis. A therapist and patient examine a specific target event in detail, identifying relevant vulnerabilities, a prompting event and successive links involving thoughts, emotions, bodily sensations, actions and environmental responses. The analysis then considers immediate and delayed consequences. Its purpose is functional: to identify how the behaviour developed in that instance and where a different response might have changed the sequence.
A chain analysis differs from a general discussion of why someone “always” reacts in a certain way. It may reveal, for example, that an interpersonal conflict was preceded by sleep disruption, an ambiguous message, an interpretation of rejection and repeated unsuccessful attempts to suppress emotion. Different links imply different intervention possibilities. A problem of inaccurate interpretation is not identical to a skills deficit, and a skills deficit is not identical to an environment that punishes effective communication.
Solution analysis follows the chain. Potential interventions include rehearsal, problem-solving, changes in environmental contingencies, cognitive strategies and exposure to situations or emotions that have been persistently avoided. DBT distinguishes lacking a skill from being unable to perform an existing skill under particular conditions. A person may understand an interpersonal technique in a calm group session but be unable to retrieve it during intense arousal; the treatment then addresses generalisation and performance, rather than simply teaching the same material again.
Validation operates alongside these change strategies. It involves identifying what is accurate, understandable or contextually meaningful in another person’s experience. A therapist may validate fear after a history of threatening interactions without agreeing that every current interaction is dangerous. Similarly, acknowledging that a harmful behaviour produced short-term relief does not endorse that behaviour. DBT describes increasingly complex forms of validation, from attentive listening and accurate reflection to recognising the validity of a response in its present context.
Dialectical strategies address rigid positions and relationship impasses. Therapists may highlight both sides of a conflict, examine apparent contradictions or shift between acceptance and change when one emphasis becomes ineffective. Commitment strategies revisit how treatment goals connect with the patient’s own priorities. Contingency management concerns the consequences that strengthen or weaken behaviour, including consequences arising within therapy. Because these procedures can be misapplied, their clinical meaning depends on transparent rationale, collaboration and attention to patient safety. The model also examines therapist behaviour: consultation is intended to identify judgement, avoidance or inconsistent responses that may interfere with treatment.
Clinical evidence and its interpretation
The most developed adult evidence base concerns people with borderline personality disorder who experience recurrent self-harm or suicidal behaviour. Trials have examined outcomes including suicide attempts, non-suicidal self-injury, hospital use, treatment retention and symptom severity. These outcomes are related but not interchangeable. A reduction in self-injury does not necessarily demonstrate fewer suicide attempts, and fewer hospital admissions do not alone establish improved everyday functioning or reduced need for care.
A 2006 trial by Linehan and colleagues compared DBT with treatment delivered by community psychotherapy experts in 101 women with borderline personality disorder and recent suicidal or self-injurious behaviour. DBT showed advantages on several suicide-related and service-use outcomes. This comparison was important because expert treatment provided a more demanding test than an unspecified usual-care condition. Nevertheless, its population and service context limit assumptions about identical results in other groups.
Not every comparison has shown DBT to be superior. A 2009 trial by Shelley McMain and colleagues compared DBT with general psychiatric management in 180 people with borderline personality disorder. Both groups improved substantially, without significant between-group differences on the principal outcomes. General psychiatric management was itself a structured, diagnosis-informed treatment rather than an absence of care. Such findings support the importance of well-organised specialist treatment while limiting claims that DBT is uniquely effective.
A 2015 component trial involving 99 women compared standard DBT with conditions that varied the inclusion of skills training and individual therapy. All conditions incorporated DBT-oriented suicide-risk management. Results supported a contribution from skills training to several outcomes, but did not establish that every patient needs exactly the same treatment package. Component studies must be read in light of the substantial clinical support shared across their comparison groups.
Systematic reviews generally support beneficial effects of DBT for some clinically important outcomes, while noting variation in study quality, comparison treatments and certainty. Psychotherapy trials cannot usually blind participants or therapists, although outcome assessors can sometimes be blinded. Small samples, attrition, investigator allegiance and inconsistent definitions of self-harm complicate synthesis. Findings against treatment as usual may differ from findings against an active specialist comparator.
The evidence is stronger for reducing certain nonfatal behaviours than for reducing suicide deaths, which are uncommon enough to require much larger samples or longer follow-up. Outcomes such as employment, durable social functioning and health-related quality of life are also important but less consistently measured. The United Kingdom’s 2009 NICE guideline on borderline personality disorder specifically identified comprehensive DBT for consideration when reducing recurrent self-harm is a priority in women. That recommendation reflects a defined population and outcome rather than a universal ranking of psychotherapies.
Adaptations across ages and clinical populations
DBT for adolescents, commonly called DBT-A, adapts the treatment to developmental needs and the young person’s social environment. Associated particularly with the work of Alec Miller, Jill Rathus and colleagues, it often includes family participation, multifamily skills groups and a “walking the middle path” module. This module addresses polarised interactions, validation and behavioural principles within families. Caregiver involvement changes the treatment system: difficulties may concern not only an individual’s skills but also recurring patterns of communication and response at home.
Trials have provided evidence for adolescent populations with repeated self-harm or elevated suicide risk. A Norwegian trial led by Lars Mehlum, reported in 2014, compared adolescent DBT with enhanced usual care. A 2018 multisite trial led by Elizabeth McCauley enrolled 173 adolescents at high risk of suicide and compared DBT with individual and group supportive therapy. DBT showed advantages on specified self-harm outcomes at the end of treatment. Differences across later assessments and outcomes illustrate why immediate benefits and their durability require separate examination.
DBT has also been adapted for substance use disorders, particularly when substance-related problems coexist with severe emotion dysregulation or borderline personality disorder. These adaptations add strategies addressing substance-related cues, attachment to treatment and responses to lapses. “Dialectical abstinence” combines an abstinence goal with a non-abandoning response if use recurs. The relevant research base is smaller than that for standard DBT, and results cannot automatically be generalised across substances, levels of dependence or service settings.
Adaptations for eating disorders have focused especially on binge eating and bulimic behaviours, where episodes may be associated with intense emotion and attempts to regulate distress. Some studies support improvements in binge-related outcomes, but programmes differ in whether they provide comprehensive DBT or selected skills. Evidence for one eating-disorder presentation does not establish effectiveness for all others, and a psychotherapy model does not replace assessment of medical complications associated with an eating disorder.
Trauma-focused developments include DBT prolonged exposure and DBT-PTSD for people with post-traumatic stress disorder and complex clinical difficulties. These are distinct protocols with their own treatment procedures and evidence. DBT prolonged exposure adds a structured trauma-focused component to DBT; DBT-PTSD integrates DBT principles with other trauma-treatment methods. Standard DBT should therefore not be assumed to constitute trauma-focused treatment simply because trauma is discussed.
Other programmes use DBT principles in inpatient, forensic, school or community settings, or adapt materials for particular cognitive and communication needs. Radically open DBT is a separately developed treatment focused on maladaptive overcontrol, rather than simply a shortened version of standard DBT. Across adaptations, the clinically relevant questions concern the target population, components actually delivered and outcomes directly studied—not the presence of the DBT name alone.
Implementation, access and unresolved research questions
Delivering comprehensive DBT requires coordinated staffing, protected consultation time, skills-training capacity and arrangements for between-session contact. These requirements affect availability within a health system. A service may offer a DBT-informed group while lacking the resources for individual therapy or coaching. Such provision can be meaningfully described and evaluated, but the label “DBT” alone does not communicate treatment intensity, clinician preparation or fidelity to the comprehensive model.
Training involves more than familiarity with skills handouts. Competence includes behavioural assessment, treatment prioritisation, management of suicidal behaviour, validation and the use of consultation. Adherence measures assess whether sessions reflect the model, although adherence and clinical skill are related rather than identical. Therapist consultation is intended both to sustain effective practice and to address difficulties generated by demanding clinical work. Its presence also makes DBT an organisational intervention to some extent, not just a technique delivered by an isolated practitioner.
Accessibility raises questions about language, disability, culture and material circumstances. Skills materials may need changes in reading level, examples, pacing or communication format. Cultural meanings of emotional expression, family obligation and interpersonal assertion can affect how a technique is understood. At the same time, adaptation can alter the intervention, making it important to document what changed and assess the result. Attendance demands, transport, work schedules and caregiving responsibilities may influence who can complete treatment and how representative trial participants are of those needing services.
Remote delivery can reduce travel barriers and support continuity, but it changes privacy, group interaction and crisis-response arrangements. Evidence from in-person comprehensive DBT cannot simply be transferred to an app, an asynchronous course or a brief online skills programme. Digital tools can record diary cards or support practice, yet their effectiveness and data privacy characteristics require separate evaluation. The relevant unit of assessment is the actual service or product, not its claimed association with an established therapy.
Economic studies examine whether treatment costs are offset by changes in emergency attendance, inpatient care and other service use. Findings depend heavily on the comparison service, local costs and follow-up period. Reduced hospital use may represent an important benefit, but economic evaluation also needs to consider functioning and quality of life. A less expensive programme is not necessarily equivalent, while a more intensive programme may not provide additional benefit for every population.
Remaining research questions include which patients benefit most from comprehensive treatment, which components are essential under different circumstances and how gains are maintained after treatment ends. More consistent reporting of adverse events, deterioration, dropout and patient-defined outcomes would improve interpretation. Trials comparing credible active treatments, examining underserved populations and following participants into ordinary social and occupational life address uncertainties not resolved by short-term changes in symptom scores.