What Is Compassion Focused Therapy and How Does It Work?

Compassion Focused Therapy is a psychotherapy model built around the idea that many psychological problems stem from an overactive threat system and an underdeveloped capacity for self-soothing. Developed by clinical psychologist Paul Gilbert, CFT draws on evolutionary psychology, neuroscience, and attachment theory to help people cultivate compassion toward themselves and others as a way of regulating difficult emotions. It was originally designed for people stuck in cycles of shame and self-criticism, and the research base has since expanded into eating disorders, trauma, psychosis, and chronic pain.

The Three-System Model Behind CFT

At the heart of CFT is a simple framework for understanding emotional life. Gilbert proposed that humans have three broad emotion-regulation systems, each shaped by evolution for a different purpose: a threat-detection system, a drive system, and a soothing system.¹ The threat system is the one most people know well. It generates anxiety, anger, and disgust to keep you safe from danger. The drive system motivates you to pursue goals, seek rewards, and compete. The soothing system, sometimes called the affiliative or safeness system, is tied to feelings of calm, connection, and contentedness. It evolved in the context of caregiving and attachment.

The core idea in CFT is that many mental health difficulties arise when these three systems are out of balance. Someone with a history of harsh criticism or neglect, for instance, may have a threat system that fires constantly and a soothing system that barely activates at all. They might also lean heavily on their drive system, pushing themselves relentlessly and then feeling inadequate when they fall short. CFT aims to strengthen the soothing system so it can act as a counterweight to threat and drive, rather than being drowned out by them.

What CFT Looks Like in Practice

CFT is typically delivered as individual therapy, group therapy, or a combination of both, often over twelve to twenty sessions. The therapist begins with psychoeducation about the three emotion systems, normalizing the client’s experience by framing their difficulties as understandable products of evolution and life history rather than personal failings. This “it’s not your fault” framing is a deliberate therapeutic move: it reduces shame before the deeper work begins.

From there, CFT uses a range of experiential exercises. One of the most distinctive is compassion-focused imagery, a guided visualization in which the person imagines a being (a person, an animal, or even an abstract figure) that radiates warmth, wisdom, and caring toward them. The aim is to generate feelings of safety and connection at a bodily level, not just as an intellectual idea.² Other exercises include compassionate letter writing, where a person writes to themselves from the perspective of a wise, compassionate part of themselves; attention training to notice self-critical thoughts without being swept up in them; and work on developing what Gilbert calls a “compassionate self,” a practiced internal stance of strength, warmth, and understanding.

CFT also draws on behavioral experiments and formulation-driven case conceptualization, sharing some DNA with cognitive behavioral therapy. But the emphasis is different. Where CBT might ask “is this thought accurate?”, CFT is more likely to ask “what would a compassionate response to this situation sound like?” The goal is not primarily to correct distorted thinking but to shift the emotional tone of the person’s relationship with themselves.

The Evidence on Shame and Self-Criticism

CFT was originally designed to help people who are deeply self-critical, and this is where the evidence is strongest. A meta-analysis in the British Journal of Clinical Psychology examined CFT’s effectiveness specifically at reducing self-criticism and increasing self-soothing, confirming that these are the therapy’s primary targets.³ A systematic review published in Behavioral Sciences drilled into the details: across eleven studies using a standard measure of self-criticism, most found significant reductions in feelings of inadequacy and self-hatred, with effect sizes ranging from small to large. The same studies found increases in self-reassurance, the capacity to comfort and encourage oneself.⁴

Shame outcomes were similarly encouraging. External shame, the feeling that others look down on you, showed significant reductions across all six studies that measured it, with moderate to large effect sizes. Follow-up data showed these reductions were maintained or even continued improving months after treatment ended.⁵ The authors noted, however, that most of these studies were of moderate quality, which means the findings are promising but not bulletproof. Whether the improvements in self-criticism hold up long-term is still unclear: one study found large gains at follow-up while another found the effects faded somewhat over time.

Conditions Where CFT Has Been Tested

A broader systematic review and meta-analysis in the Journal of Affective Disorders looked at CFT across multiple clinical populations and found it effective at improving compassion-based outcomes and reducing symptoms from baseline to post-treatment, and compared to waitlist controls. Effect sizes ranged from small to large across self-compassion, self-criticism, self-reassurance, depression, and eating disorder symptoms.⁶ That range is worth noting: CFT does not produce uniformly dramatic effects, but it reliably moves the needle in the right direction across quite different problems.

Trauma and Complex PTSD

Shame and self-blame are central features of trauma-related disorders, making CFT a natural fit. In a multiple-baseline study of ten participants receiving compassion-based therapy for trauma-related shame, nine showed reliable decreases in PTSD severity and eight showed reliable reductions in shame. These gains held at follow-up, and participants reported high satisfaction.⁷ A case study of complex PTSD published in the Journal of Clinical Psychology described clinically significant reductions in CPTSD symptoms and trauma-related shame over the course of CFT treatment, alongside improved emotional regulation and a stronger compassionate identity.⁸ CFT has been described as a transdiagnostic approach that helps individuals shift from shaming and self-critical emotional responses to patterns oriented around understanding and wisdom, which is particularly relevant for CPTSD, where disturbances in self-organization are a defining feature.⁹

Eating Disorders

Shame about one’s body and eating behaviors is pervasive in eating disorders, and CFT has been applied in this area with growing interest. A systematic review in the Journal of Eating Disorders found that CFT led to reductions in core eating disorder symptoms, improvements in self-compassion and body image, and decreases in shame.¹⁰ A separate scoping study found that CFT-combined approaches were particularly helpful for reducing body image dissatisfaction and eating pathology in young adult females with binge eating disorder.¹¹

Psychosis and Voice-Hearing

CFT has been adapted for people who hear distressing voices, an application that might surprise those who associate compassion work only with mood disorders. The rationale is that voice-hearers often experience their voices as dominant and shaming, recreating a kind of internal bullying dynamic. CFT aims to help them notice these threat-based patterns and shift into emotional states oriented around safeness.¹² A case series of seven participants with psychosis found significant improvements in depression, stress, and distress related to voices and delusions. The improvements in voices and delusions were sustained at follow-up, though gains in depression and stress dropped slightly.¹³

Chronic Pain

People living with persistent pain often fall into cycles of striving (pushing through pain, then crashing) and self-blame (feeling like a failure for not coping better). A study applying CFT’s three-systems approach to chronic pain found significant improvements in depression, self-compassion, pain-related disability, pain-related anxiety, and pain self-efficacy.¹⁴ The logic here is not that compassion cures pain, but that reducing the self-critical overlay on pain frees up a person’s capacity to manage their condition more effectively.

How CFT Compares to CBT

Given that CFT grew partly out of cognitive behavioral therapy, the obvious question is whether one works better than the other. A randomized controlled trial compared CFT to CBT for adults with eating disorders, some of whom had a history of childhood trauma. Both treatments produced significant improvement over time. But among patients who had experienced trauma, CFT showed a more substantial impact on eating disorder symptoms during the follow-up period, suggesting it had a particular advantage for trauma-affected individuals that became more apparent after treatment ended.¹⁵

This finding fits CFT’s theoretical rationale. CBT excels at identifying and restructuring unhelpful thought patterns, but for people whose core difficulty is not so much distorted thinking as an impoverished capacity for self-warmth, CBT’s tools may not reach the underlying problem. A person can learn that their thought “I’m worthless” is irrational and still feel it in their bones. CFT targets that felt sense directly by training the soothing system rather than focusing on the accuracy of thoughts. In practice, many therapists draw on both approaches, using cognitive restructuring alongside compassion-based exercises.

What Happens in the Body

CFT is not just talk therapy; it makes claims about what is happening physiologically. The theoretical link is through the vagus nerve and the parasympathetic nervous system, which governs the body’s “rest and digest” state. Compassion-based approaches are thought to activate vagal tone, which can be measured through heart rate variability (HRV), a biomarker of the body’s capacity to flexibly shift between physiological states.¹⁶

The evidence here is nuanced. A 12-week group CFT intervention found that resting HRV did not change significantly for the whole sample. However, participants who showed a reliable increase in self-compassion also had significantly increased resting HRV. The study also found that self-critical writing caused HRV to drop, and that people who gained more self-compassion through CFT showed greater HRV responsiveness during emotional writing tasks afterward, suggesting that the therapy may strengthen the body’s ability to engage with difficult emotions rather than shut down.¹⁷ A separate study of CFT for overeating found that improvements in self-compassion correlated with positive changes in HRV, consistent with the idea that compassion practice modulates the vagal system.¹⁸

Brain imaging studies add another layer. A scoping review of neuroimaging research on compassion-based interventions found that these practices were consistently associated with increased activity in prefrontal and mesolimbic brain regions and altered activity in posterior parietal and occipital areas. Some studies also found increased connectivity between frontal and reward-related areas of the brain.¹⁹ An earlier neuroimaging study found that self-criticism activated the anterior insula, anterior cingulate cortex, and amygdala (regions associated with threat processing) significantly more than self-reassurance did, offering a neurological basis for why self-critical people live in a state of heightened threat.²⁰ One notable gap: an early study investigating whether compassion training changes plasma oxytocin found that overall oxytocin changes were unrelated to cortisol and subjective stress responses, which complicates the popular narrative that compassion simply “boosts oxytocin.”²¹

When Compassion Feels Threatening

One of CFT’s most distinctive contributions to the therapy field is the recognition that compassion itself can feel dangerous. For people who grew up in environments where vulnerability was punished or warmth was unpredictable, turning compassion inward can trigger anxiety, grief, or even panic. Gilbert and colleagues developed the Fears of Compassion Scales to measure this phenomenon and found that it runs along three channels: fear of being compassionate toward yourself, fear of being compassionate toward others, and fear of receiving compassion from others.²²

A meta-analysis of the research on fears of compassion found that these fears are meaningfully linked to mental health difficulties. The overall association between fear of self-compassion and mental health problems was moderate to strong, with the strongest links to shame, self-criticism, and depression. Fear of receiving compassion from others showed similarly strong associations.²³ This is clinically important because it means CFT therapists cannot simply instruct a client to be kinder to themselves and expect the soothing system to activate. For many clients, the early work in CFT involves gradually making it safe to receive compassion at all, often at a pace that surprises therapists trained in other modalities.

Attachment style plays a role here. A study using physiological measures found that individuals with higher levels of anxious or avoidant attachment were more likely to show a threat response, measured as a decrease in heart rate variability, during their first exposure to compassion-focused imagery.²⁴ In other words, the very people who most need the soothing system to come online are often the ones whose bodies react to compassion cues as if they were threats. CFT addresses this directly, spending considerable time helping clients understand why compassion feels unsafe and building tolerance gradually.

Cultural Adaptation

CFT was developed in a Western context, and its assumptions about what healthy self-relating looks like do not always translate seamlessly. Research has highlighted significant differences between East Asian and Western populations in CFT-relevant variables: lower self-compassion, higher self-criticism, higher external shame, and different structures of fears of compassion in East Asian samples. A detailed account of adapting CFT for Japanese clients involved moving from a sixteen-session UK model to a twelve-session culturally refined version. The adaptations went beyond translation. Therapeutic goals in the East Asian context involved a more moderated form of self-expression, balancing internal compassionate self-acceptance with external humility. Clinicians found they needed to validate self-criticism as a coping style rather than immediately trying to reduce it, introduce case formulation earlier to address fears of compassion non-judgmentally, and prioritize compassionate-other imagery before moving to compassionate-self imagery.²⁵ The adapted program for treatment-resistant depression showed high effectiveness, though the authors emphasized that meaningful cultural adaptation is challenging and requires ongoing clinical feedback rather than a one-time translation effort.

CFT for Adolescents in Care

One area where CFT is being cautiously explored is with young people who have experienced early adversity. A group program was developed for adolescent girls aged twelve to sixteen living in foster, adoptive, or kinship care, framed as a “Building Your Self-Confidence” group delivered over eight weekly sessions. Most participants attended all sessions and found the group enjoyable and interesting. Half of the caregivers reported improvements in their child’s mental health on standardized measures, though none of the adolescents themselves reported improvement on those same measures.²⁶ That discrepancy between caregiver and adolescent self-report is common in youth mental health research and does not necessarily mean the intervention failed, but it highlights the difficulty of measuring change in a population that may distrust authority figures and self-report instruments alike. Work with young people in care represents an early frontier for CFT, and the evidence so far is preliminary rather than established.

The Therapist’s Own Compassion Practice

An underappreciated aspect of CFT is its emphasis on the therapist’s own relationship with compassion. CFT therapists are expected to practice what they teach, cultivating self-compassion not just as a professional tool but as a personal practice. Research on mental health professionals suggests this is not just aspirational: one study found that changes in self-compassion scores predicted a substantial portion of the variance in burnout reduction among clinicians, with self-compassion training accounting for roughly a third of the change in burnout outcomes.²⁷ The implication is that a therapist’s own capacity for self-compassion is not just personally beneficial but may affect the quality of the therapeutic relationship and the credibility of the compassion-based interventions they deliver. A therapist who teaches compassion while running on fumes and self-criticism is, in CFT’s framework, modeling exactly the imbalance the therapy is designed to correct.