CBT for Psychosis: Goals, Evidence, and Limitations
Author: Ukrainian Psychological Hub · Published: October 9, 2026 · Editorial Policy
Cognitive behavioral therapy for psychosis (CBTp) is a collaborative psychological treatment that helps people understand distressing psychotic experiences, experiment safely with different responses, and pursue goals that matter in everyday life. It may address hearing voices, fears of being watched or harmed, social withdrawal, troubling interpretations of events, or loss of confidence after an episode. Its goals can include less distress, greater flexibility, improved relationships, and a return to education, work, or ordinary activities. A reduction in hallucinations or delusions may occur, but it is not the only meaningful definition of progress.
CBTp is recommended in major schizophrenia treatment guidelines, and research supports modest average benefits for some symptom and functioning outcomes when added to standard care. Evidence is less conclusive for preventing relapse, outperforming other active psychological interventions, or replacing medication. An accurate account of CBTp therefore distinguishes the goals of therapy from the effects established in clinical trials. Sources: APA practice guideline, NICE CG178, and 2024 Cochrane review.
This guide covers CBT specifically for psychotic experiences. The broader syndrome is explained in Psychosis: Symptoms, Causes, Types, and When to Seek Help, while the comprehensive diagnosis and treatment pathway for schizophrenia is covered in Schizophrenia: Symptoms, Causes, Diagnosis, Treatment, and Recovery.
What is CBT for psychosis?
CBTp uses ideas and methods from cognitive behavioral therapy but adapts them to the unusual perceptions, strongly held beliefs, emotions, coping responses, social consequences, and clinical needs associated with psychosis. It begins with an individualized formulation: a shared working explanation of the circumstances in which a problem occurs, how the person interprets it, what they feel, how they respond, and which consequences might maintain fear or disruption. The formulation remains open to revision in response to the person's experiences and new information.
For instance, someone who hears a threatening voice may fear that disobeying it will cause harm. Treatment can explore how perceived power, predictions, stress, avoidance, and coping affect the distress. A voice might become less commanding even if it remains audible. Someone with suspiciousness might begin traveling independently or contacting friends while still having periods of uncertainty. CBTp can focus on changes in agency and functioning as well as symptom measures.
CBTp does not require a therapist to argue someone out of their experiences or insist that the person adopt the clinician's interpretation. A competent approach involves curious questions, safe collaborative observations, attention to evidence, and agreement on meaningful goals. The NICE guideline explicitly frames CBT as helping a person connect thoughts, feelings, behavior, symptoms, and functioning, while encouraging alternative coping and reduced distress.
Who is CBTp for? Psychosis, schizophrenia, and diagnostic context
Psychosis describes a clinical presentation that may involve hallucinations, delusions, disorganized thinking, and changes in behavior. Schizophrenia is one diagnosed disorder associated with psychosis; related conditions include schizoaffective disorder, schizophreniform disorder, brief psychotic disorder, and delusional disorder. Psychotic symptoms may also appear during bipolar mood episodes, major depression with psychotic features, substance intoxication or withdrawal, and neurological or other medical illnesses. These distinctions determine the broader treatment plan.
One unusual experience, a screening score, or a brief description of paranoia cannot establish a disorder. Assessment considers onset, duration, mood episodes, medical history, prescribed and nonprescribed substances, changes in attention and cognition, sleep, safety, and functional impairment. New confusion, fluctuating awareness, fever, seizures, intoxication, or neurological symptoms require prompt medical assessment rather than a routine outpatient therapy exercise. For context see Schizophrenia Spectrum Disorders, Psychosis Due to a Medical Condition, and Schizophrenia vs. Bipolar Disorder.
CBTp research has often included people with schizophrenia-spectrum diagnoses. Its findings should be applied thoughtfully to other clinical contexts. For psychosis related to a manic episode, severe depression, or an acute medical problem, the cause-specific pathway remains central. CBTp may be useful alongside other care when a qualified clinician assesses it as appropriate; the symptom itself is not a complete treatment plan.
What are the main goals of CBTp?
Reducing distress rather than pursuing symptom scores alone
A person may choose to feel less frightened by voices, reduce time spent checking for danger, sleep better, regain the ability to use public transportation, or reconnect with a family member. The experience might not disappear immediately. The therapist evaluates how much distress it produces, what it stops the person from doing, and whether the person can regain choice in responding. Improving life with persisting experiences can be a clinically valuable outcome.
Exploring interpretations and threat expectations
The meaning assigned to an unusual perception can be crucial to how distressing it becomes. A sound interpreted as evidence of surveillance may trigger panic and withdrawal; the same sound considered uncertain may be less controlling. The therapist can help compare explanations, examine predictions, notice factors such as poor sleep, and build tolerance for uncertainty. The aim is a more workable understanding built collaboratively rather than an imposed declaration about a belief.
Reducing avoidance and unnecessary safety behaviors
Repeated checking, reassurance seeking, staying indoors, or monitoring other people may temporarily reduce anxiety while maintaining an expectation of danger. If a person wishes, the therapist can help identify a safer, more flexible alternative and evaluate what happens. Real hazards, harassment, discrimination, and abuse must be taken seriously. CBTp never requires exposure to an objectively unsafe circumstance as a supposed test of a belief.
Restoring meaningful activity and social participation
Goals often extend beyond voices and beliefs. A person may wish to return to school, sustain work, handle a daily schedule, participate in family life, or enjoy hobbies. Therapy can work on hopelessness, shame, avoidance, and confidence in small, cumulative steps. Supported employment, family services, physical health care, medication review, and rehabilitation may be needed alongside psychotherapy to make these goals attainable. APA guideline.
How CBTp works in practice
Engagement, consent, and a shared agenda
Early appointments focus on listening, understanding what the person wants, explaining confidentiality and its limits, and establishing a respectful working relationship. The therapist asks which experiences are most intrusive and which changes would be useful. Some people want immediate coping strategies; others first need to be heard or to understand what has happened during a frightening clinical episode. Agreement on the target problem is a foundation of collaborative treatment.
Building a personal formulation
A formulation may map a trigger, the experience, its meaning, emotional and bodily responses, coping behavior, and later consequences. For example, after several nights of poor sleep, ambiguous noises may feel especially threatening; fear increases vigilance; constant checking makes sleep worse; and withdrawal reduces opportunities to encounter reassuring evidence. The model is a tentative account of a feedback loop, not a claim that thoughts cause every symptom of psychosis.
Guided discovery and behavioral learning
The clinician and client can explore evidence for several interpretations, review predictions, and design small experiments that fit the client's priorities. A safe experiment might involve spending a short period in a familiar environment without a particular checking ritual and noticing what happens, provided it creates no actual danger. The exercise is agreed upon, never coerced, and reviewed afterward. Distress, medication needs, and acute medical warning signs remain clinically relevant throughout.
Coping strategies and attention to daily conditions
CBTp can include activity planning, strategies for disengaging from distressing voices, attention-shifting, reducing overwhelming stimulation, improving sleep routines, asking for support, and organizing challenging tasks. These techniques are adapted to cognitive capacity and daily circumstances. They work best when connected to a specific formulation and measurable personal goal rather than presented as a universal collection of self-help tricks.
Review, consolidation, and planning for setbacks
A course of therapy revisits goals and outcomes, reviews what was helpful, identifies early signs of escalating stress, and clarifies how to contact the treatment team. A personalized care plan may include routines, trusted supports, and agreed procedures for renewed distress. A recurrence of symptoms calls for assessment and responsive care; it does not erase progress already made. NICE monitoring recommendations.
What happens during a CBTp session, and how many sessions are needed?
Appointments commonly begin with a check-in about the past week, including sleep, stress, safety, and important events. Therapist and client review a previous task, choose a problem to explore, identify its interpretation and consequences, practice a way of responding, and agree on a manageable observation or activity before the next meeting. Some sessions are primarily about trust, difficult emotions, or practical planning; a rigid script may not fit everyone.
For the specific individual protocol described in its adult guideline, NICE recommends at least 16 planned one-to-one sessions delivered with a manual-informed approach, including monitoring or coping work and attention to distress or functioning. Access, pace, frequency, and personal capacity vary. Brief CBTp-informed conversations, group programs, and a complete individual course should be described as distinct interventions rather than assumed to have identical evidence. NICE CG178, section 1.3.7.
CBTp for hearing voices and hallucinations
Therapy for voice-hearing explores the timing, content, perceived identity or authority, and practical impact of voices. A person might fear that a voice can punish them, interpret a voice as omnipotent, or spend much of the day obeying or arguing with it. The work may reduce fear, build alternative coping responses, and strengthen the ability to choose which actions to take. The person's cultural understanding and preferences belong in the conversation.
Some treatments focus on beliefs about the voice's power and the perceived consequences of resisting commands. The COMMAND randomized trial tested a specialized cognitive approach for people with persistent command hallucinations and a history of harmful compliance. Results from that selected population cannot be generalized to everyone who hears voices, and this technique is not necessarily part of every CBTp program.
Clinical assessment is essential when voices encourage suicide, self-injury, or violence. The clinician considers perceived compulsion, intention, past actions, access to means, protective factors, and immediate risk. A command hallucination by itself is not evidence that the person will act; when there is immediate danger or the person feels unable to resist a dangerous command, emergency clinical care takes priority over a therapy exercise.
CBTp for paranoia and persecutory delusions
A person may believe that others are watching, plotting against, or trying to harm them. CBTp can examine the experience of threat, ambiguous social cues, hypervigilance, sleep, reasoning, checking, and avoidance. Therapists should remain sensitive to real interpersonal threats and discrimination. They explore interpretations while respecting that safety is a practical question requiring attention to actual circumstances.
In the Feeling Safe randomized controlled trial, researchers tested a specialized cognitive treatment against befriending for people with persistent persecutory delusions. It supports the possibility of targeted change in such beliefs and related distress. Its findings apply to that protocol and study population, and should not be presented as a guaranteed response to a generic appointment described simply as CBT.
A graded experiment may involve entering a familiar safe space, clearly specifying the feared outcome, reducing an unnecessary protective ritual, and observing what occurs. The client retains control over the pace and can stop or adapt a plan. CBTp must never require abandoning precautions against credible danger or deliberately confronting a potentially violent person.
Negative symptoms, cognitive impairment, and everyday functioning
Psychotic disorders may involve reduced motivation, limited speech or emotional expression, diminished pleasure, withdrawal, difficulties with attention, memory, or planning, and practical barriers to work or education. CBTp may help where hopelessness, distress, avoidance, or expectations maintain inaction. Other causes, including depression, sedation, adverse drug effects, poverty, and illness-related cognitive changes, may call for different or additional interventions.
When attention and executive functioning are impaired, psychotherapy may use written reminders, simplified tasks, repetition, shorter exercises, and concrete goals. Cognitive remediation directly targets cognitive functions and learning strategies, while family psychoeducation and supported employment address other aspects of recovery. CBTp works within an integrated plan and should not be represented as a substitute for all rehabilitation services. VA/DoD guideline.
What do major clinical guidelines recommend?
The American Psychiatric Association schizophrenia practice guideline, 2020 recommends that people with schizophrenia receive CBTp (recommendation 16, grade 1B). The NICE adult guideline CG178 recommends offering individual CBT to people with psychosis or schizophrenia, including during acute or later care and in appropriate inpatient settings. The 2023 VA/DoD guideline suggests CBTp alongside pharmacotherapy for relevant early-psychosis populations and lists CBTp among psychosocial treatments for schizophrenia.
Guidelines consider effectiveness, safety, availability, values, and the place of an intervention in comprehensive care. A recommendation to offer CBTp does not say that it works equally well for every person or that it is the superior therapy for every measurable outcome. The randomized evidence provides a more specific view of expected benefit, which is essential for informed choice.
Does CBTp work? A close reading of the scientific evidence
Overall and positive psychotic symptoms
A 2018 network meta-analysis by Bighelli and colleagues analyzed 53 randomized controlled trials involving 4,068 people with schizophrenia-spectrum conditions and positive psychotic symptoms. CBT contributed 40 trials; relative to treatment as usual, the estimated improvement in positive symptoms was a standardized mean difference of approximately −0.30 (95% confidence interval −0.45 to −0.14). This means a small average difference on a statistical scale, not a 30% cure rate. Effects and certainty varied with the comparator and trial design.
The 2014 review by Jauhar and colleagues highlighted methodological concerns, including smaller estimated effects in studies with blinded assessments. Because psychotherapy trials cannot usually blind participants or therapists to the therapy received, independent outcome assessment and active comparison groups are particularly important. A large effect reported by an uncontrolled program should not be treated as equivalent to evidence from carefully conducted randomized trials.
First-episode and recent-onset psychosis
The 2024 Cochrane review by Mayer and colleagues included 28 trials, with usable data from 26 trials involving 2,407 participants. When added to standard care, CBT produced modest pooled differences favoring treatment for overall symptoms (SMD −0.27), positive symptoms (−0.22), negative symptoms (−0.20), and functioning (−0.23), mostly within approximately one year. Certainty of the evidence ranged from moderate to very low depending on the outcome. Individual results can differ substantially from those averages.
The same review did not establish a clear difference in relapse: relative risk 0.82 with a 95% confidence interval from 0.57 to 1.18, a range including no improvement. It also had very limited information about adverse events. Benefits in symptom severity or functioning therefore should not be translated into a proven promise to avoid hospitalization, relapse, or psychological harms. Cochrane review details.
Is CBTp better than other psychological treatments?
A 2018 Cochrane review of CBT versus other psychosocial treatments included 36 randomized trials and 3,542 participants. It did not identify clear convincing superiority for several important outcomes, including relapse, readmission, symptoms, and quality of life; much evidence was low or very-low certainty. Another Cochrane review of CBT added to usual care included 60 trials and did not clearly establish a long-term relapse reduction. Comparisons against supportive, active therapy can produce different conclusions than comparisons against minimal care.
What about group CBTp?
A Cochrane review of group CBTp considered 24 studies with about 1,900 participants. It suggested potential improvements in total symptoms but found uncertain or absent clear differences on several specific symptom domains. Group therapy, individual manualized CBTp, targeted paranoia programs, and low-intensity CBTp-informed care are different intervention formats. Their effects should be reported using the evidence for each format and population.
What research limitations matter most?
The meaning of a trial's result depends on the people enrolled, severity of illness, additional medication and services, therapeutic relationship, competence and supervision, intervention dose, controls, and follow-up length. Research outcomes may emphasize symptom rating scales while readers care more about sleep, social connection, autonomy, or continuing education. Differences in outcome measures, missing data, and limited blinding make some findings uncertain. The appropriate scientific conclusion is that CBTp is a useful evidence-supported option with modest mean benefits and significant unresolved questions.
Can CBTp replace antipsychotic medication?
Psychosis care commonly combines diagnostic assessment, antipsychotic medication when indicated, psychotherapy, and psychosocial support. The treatment components address different needs. Medication decisions must involve a prescriber, discussion of potential benefits and adverse effects, physical-health monitoring, and shared decisions responsive to the person's preferences and history. CBTp should not become a reason to abruptly stop or independently change prescribed medication.
A 2024 Cochrane review of CBT without antipsychotics included only four trials and 300 participants and found low- or very-low-certainty evidence. About 35% of participants assigned to CBT without medication received antipsychotic treatment during the studies, further limiting interpretation. In some analyses CBT without medication produced less improvement in positive symptoms than CBT combined with medication. The review could not support strong conclusions about medication-free CBT as a reliable alternative.
The NICE first-episode guidance recommends offering oral antipsychotic medication alongside psychological interventions. It also describes a monitored shared-care option when a person chooses to try psychological interventions alone, including early review of treatment choices. That discussion is best handled individually, in light of the diagnosis, current risks, patient goals, and clinical circumstances. There is no basis for an unsupervised general recommendation to substitute CBTp for pharmacotherapy.
First-episode psychosis and clinical high-risk states
An early psychotic episode can disrupt education, employment, family relationships, and confidence at a particularly important stage of life. CBTp may be one component of coordinated specialty care, which typically combines clinical treatment, family education, supported work or education, and follow-up. Evidence from first-episode studies supports modest improvements when CBT is added to standard services. The APA guideline recommends coordinated specialty care for schizophrenia with a first episode.
A clinical high-risk state involves specialist assessment of attenuated or transient psychotic experiences and other contextual criteria; it is not a confirmed schizophrenia diagnosis and does not predict inevitable transition. NICE recommends individual CBT for appropriately assessed people at increased risk and advises against offering antipsychotics solely to prevent psychosis. Monitoring, supportive intervention, and evaluation of anxiety, depression, trauma, sleep, or substance use may be relevant to the particular presentation.
CBTp as part of recovery-oriented care
A recovery-oriented model asks how care can improve participation, personal choice, relationships, safety, and satisfaction with life. People may need help with housing, physical health, social isolation, stigma, educational access, substance-use problems, or medication side effects. CBTp can address coping and beliefs that interfere with goals, but these wider needs require appropriate social and clinical services as well.
Clinicians should review whether apparent therapy nonresponse reflects a poor therapeutic fit, insufficient access, cognitive limitations, untreated mood symptoms, adverse effects, trauma, social adversity, or the need for additional specialist care. Family support can improve coordination when the person consents. The VA/DoD psychosocial guideline synopsis describes CBTp within a broader set of interventions rather than as a stand-alone solution to schizophrenia.
Potential drawbacks, safety, and limits
Talking about frightening experiences may briefly increase anxiety; behavioral experiments can be uncomfortable; and some people have experienced coercive or stigmatizing encounters with services. Treatment should explain goals and possible burdens, adapt to culture and trauma history, and allow feedback or changes in direction. Trials have incompletely measured psychotherapy-related adverse effects, so the limited available evidence should not be interpreted as proof that no harm can occur.
CBTp cannot confirm or exclude schizophrenia through a conversation, reliably prevent every relapse, eliminate all voices, or resolve psychosis caused by delirium, intoxication, or a serious neurological disorder. Its appropriate role depends on formulation, professional training, acute safety, and shared treatment planning. A person whose therapy is not helping deserves a careful reassessment of goals, access, diagnosis, and alternative or complementary treatments.
How to find qualified CBTp treatment
A psychiatrist, early-psychosis service, community mental health center, academic training clinic, or specialized psychological treatment program may know clinicians trained in CBTp. Ask about psychosis-specific training, clinical supervision, manualized interventions, formulation, safety management, and coordination with a prescribing clinician. Availability in the United States varies by region, payment arrangements, and service capacity. The NAMI CBTp information page lists examples of training resources.
Useful questions include: What are your CBTp qualifications? How do we decide on goals? Have you treated distressing voices or persecutory beliefs? What does a complete course involve? How will we measure progress in distress and functioning? How do you handle trauma, medical causes, and urgent risk? How can care be coordinated with medication management? If a clinic offers only generic CBT, ask whether a psychosis-trained clinician or referral service is available.
How to measure meaningful progress
Treatment reviews can include voice-related distress, conviction about a particular threat, time spent checking, avoidance, activity levels, sleep, engagement in education or work, relationships, and the person's judgment of recovery. Standardized symptom measures may complement individually chosen goals, but they do not replace them. Less distress can be meaningful even when hallucinations persist, while fewer hallucinations need not automatically translate into better functioning.
The NICE guidance recommends routine evaluation of psychological intervention outcomes and access to care. Progress should be discussed openly rather than inferred solely from the clinician's preference for a particular explanation. Outcomes may fluctuate, and a care plan can be adjusted when goals, symptoms, or circumstances change.
When to seek urgent assessment instead of waiting for psychotherapy
Rapidly developing hallucinations or delusions accompanied by fluctuating awareness, fever, seizure-like activity, severe physical illness, new neurological symptoms, intoxication, or withdrawal need urgent medical evaluation. These presentations may have a treatable medical or substance-related cause. See the English Hub's Psychosis Due to a Medical Condition article for the separate diagnostic and safety questions.
Immediate danger, suicidal intent, a plan to harm someone, severe inability to care for basic needs, or dangerous commands that feel irresistible warrant emergency help. In the United States, 911 is appropriate for immediate danger, and 988 provides crisis support. Describe the present circumstances and known clinical history to responders. An ordinary scheduled CBTp appointment is not an adequate substitute for emergency assessment.
Frequently asked questions
Is CBTp the same as standard CBT?
It shares the cognitive and behavioral principles of standard CBT, but uses specialized formulation, engagement, and interventions for psychotic experiences and their meaning. Generic CBT for anxiety is not automatically an equivalent treatment.
Does CBTp try to prove voices or beliefs are unreal?
A collaborative therapist explores the impact of an experience and alternative ways to interpret and respond. Demanding agreement or dismissing a person's experience undermines engagement and may increase distress.
Can CBTp stop hearing voices?
Some people report less frequent or intense experiences, while others benefit primarily through reduced fear, perceived power, or disruption. A predictable disappearance of voices cannot be promised.
How many CBTp sessions should I expect?
NICE describes at least 16 planned individual sessions for its recommended protocol. Local programs and shorter CBTp-informed approaches vary; ask the provider exactly what is offered.
Does CBTp help paranoid thoughts?
Therapy can explore threat expectations, anxiety, safety behaviors, and the impact of suspicious interpretations. Specialized randomized trials support the potential of targeted approaches while leaving individual outcomes uncertain.
Can CBTp work without medication?
Research on CBTp as a replacement for antipsychotic treatment is limited and low certainty. Shared decisions and regular clinical monitoring are important when a person declines medication.
Is CBTp used before a confirmed schizophrenia diagnosis?
Specialists may offer CBT to some people meeting clinical high-risk criteria, but these criteria are not a diagnosis and do not mean psychosis or schizophrenia will inevitably develop.
Can family members deliver CBTp at home?
Families can learn supportive communication and take part in evidence-based family interventions. Formal CBTp is a specialist clinical treatment and should not require relatives to confront unusual beliefs or take responsibility for managing acute risk.
Is CBTp different from cognitive remediation?
CBTp primarily addresses the meaning, distress, and behavior associated with psychotic experiences. Cognitive remediation more directly addresses cognitive abilities and compensatory learning strategies.
What if the therapy feels unhelpful?
Discuss the fit of the formulation, relationship, goals, pace, and outcome measures with the clinician. Reassessment, a different format, additional rehabilitation, or another care approach may be appropriate.
Conclusion
CBTp is an evidence-supported component of care for psychosis that works through collaboration, personalized formulation, safe behavioral learning, and recovery-focused goals. Guidelines support offering it, and trials suggest modest average improvements in several clinical domains, particularly alongside standard treatment. Important uncertainties persist about preventing relapse, outperforming other active therapies, and substituting psychotherapy for medication. Effective care combines respect for the person's priorities with diagnostic accuracy, accessible treatment, and appropriate medical and psychosocial support.
Related Articles
Schizophrenia: Symptoms, Causes, Diagnosis, Treatment, and Recovery — the central guide to schizophrenia diagnosis, treatment, and recovery.
Psychosis: Symptoms, Causes, Types, and When to Seek Help — clinical meaning, causes, and assessment of psychosis.
Schizophrenia Spectrum Disorders: Types, Differences, and Clinical Diagnosis — spectrum diagnoses and boundaries.
Psychosis Due to a Medical Condition: Causes and Evaluation — acute medical and neurological differential.
Schizophrenia vs. Schizoaffective Disorder: Key Diagnostic Differences — the diagnostic distinctions relevant to treatment.